(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Bulletin of the British Museum (Natural History)"

/%/ r/ 



r-D 



J^S^ 



(oio 



b 8 m s. 

BULLETIN OF 
THE BRITISH MUSEUM V!u.ra«.^ 
(NATURAL HISTORY) 



BOTANY 

Vol. 4 
1965-1973 



BRITISH MUSEUM (NATURAL HISTORY) 
LONDON: 1975 



2 2MMi«5 



DATES 


OF PUBLICATION 


OF THE PARTS 


No. I .... . November 1965 


No. 2 










May 1967 


No. 3 










June 1968 


No. 4 










May 1969 


No. 5 










June 1969 


No. 6 










August 1970 


No. 7 










August 1 97 1 


No. 8 










December 1973 



ISSN 0068-2292 



Printed in Great Britain hyjohn Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 5NU 



2 2 mi ^^^5 




CONTENTS 

BOTANY VOLUME 4 

PAGE 

No. I. Cuticular studies as an aid to plant taxonomy. By C. A. Stage . i 

No. 2. The genus Elaphoglossum in the Indian Peninsula and Ceylon. By 

W. A. Sledge .......... 79 

No. 3. Fungi of recent Nepal expeditions. By F. L. Balfour-Browne . 97 

No. 4. A synopsis of Jamaican Myrsinaceae. By W. T. Stearn . . 143 

No. 5. The Jamaican species of Columnea and Alloplectus (Gesneriaceae) . 

By W. T. Stearn ......... 179 

No. 6. New or little known Himalayan species of Swertia and Veratrilla 

(Gentianaceae) . By H. Smith ....... 237 

No. 7. A survey of the tropical genera Oplonia and Psilanthele (Acantha- 

ceae). By W. T. Stearn ........ 259 

No. 8. Angiosperms of the islands of the Gulf of Guinea (Fernando Po, 

Principe, S. Tome, and Annobon). By A. W. Exell . . . 325 

Index to Volume 4 ........ . 413 



> n ' t^( . ^ • 



CUTICULAR STUDIES AS 

AN AID TO 

PLANT TAXONOMY 



;j:4l V- 



CLIVE A. STAGE 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. i 

LONDON: 1965 



CUTICULAR STUDIES AS AN AID 
TO PLANT TAXONOMY 




BY 



CLIVE A. STAGE 



(University of Manchester) V 



Pp. 1-78 ; 10 Text-figures ; Plates 1-5 



BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

BOTANY Vol. 4 No. i 

LONDON: 1965 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

This paper is Vol. 4, No. i of the Botany series. 



© Trustees of the British Museum (Natural History), 1965 



trustees of 
the british museum (natural history) 

Issued November, 1965 Price One Pound Fifteen Shillings 



CUTICULAR STUDIES AS AN AID 
TO PLANT TAXONOMY 

By CLIVE A. STACK 



This review is an account of the systematic value of the epidermal characters ol 
angiosperm leaves as they are seen on preparations of the cuticular membrane, and is 
based on a detailed study of about 250 species belonging to three families. The 
introductory sections outline the history of the taxonomic use of cuticular and epider- 
mal characters, the anatomical background of the features observed on cuticular 
membranes, the methods employed in the isolation and mounting of the latter, 
and a suggested routine for their formal description. The main part of the work 
comprises a somewhat detailed account of the cuticular characters which have been 
found to be useful in identification and taxonomy, and of the variability that some 
of these exhibit in various conditions. The precautions necessary in the use of such 
characters are stressed. In the final section the place of cuticular characters in 
modern taxonomic work is discussed. 

It is hoped that these comments will help to place the systematic use of the 
epidermis upon a firmer and more scientific basis than has hitherto been the case. 

This paper is a modified portion of a thesis submitted to the University of London 
for the Ph.D. degree. The work was supported by grants from the Department of 
Scientific and Industrial Research. 

I am indebted to the Keeper of Botany, British Museum (Natural History), for 
providing laboratory, library and herbarium facilities for the pursuit of this work, 
and to Dr. A. W. Exell for his very considerable help and valuable discussion at all 
stages. I am also grateful to those members of the Departments of Botany and 
Palaeontology, British Museum (Natural History), who have helped me with several 
problems, and to Mr. G. Grange, Department of Botany, University of Manchester, 
for preparing the photomicrographs. 

References (by author and date) given in the text are to the Ust of Special Literature 
at the end of the paper. 



4 cuticular studies as an aid to plant taxonomy 

§1, Preface 

Plant taxonomy has undergone a great many changes since it was first treated as a 
scientific disciphne in the seventeenth century, one of the major contributing factors 
being the discovery of the use of new fields of investigation to supplement the old. 
The ultimate aims of the majority of taxonomists have naturally also changed, the 
original hope of cataloguing every species of plant now embracing the idea that their 
evolutionary and genetic interrelationships should also be expressed. I accept that 
taxonomists should aim at devising " natural " classifications which attempt to 
indicate the phylogenetic position of each basic unit, although it seems to me certain 
that this goal will be in sight only when the genetical constitution of each taxon is 
known and can be expressed with something approaching the precision of a chemical 
formula, towards which the present binomial system of nomenclature is no more than 
a preliminary step. 

Prior to the latter part of the nineteenth century taxonomists used only morpholo- 
gical (i.e. macroscopic) characters in their studies, and apart from the inclusion of the 
broadest vegetative features they confined themselves to the reproductive organs. 
It was not until about 1865 that anatomical (i.e. microscopic) characters were first 
used, and considerably later that this became usual practice. The volumes of 
Engler & Prantl's Die naturlichen Pflanzenfamilien (1887-1915) included a relatively 
large amount of anatomical and vegetative morphological details, and since then 
taxonomic monographs have been considered incomplete without them. Modern 
workers are constantly employing new diagnostic characters, and work of this kind 
will need to continue until every structural, physiological and biochemical feature 
has been utihzed. 

Parts of the plant which have been most successfully investigated from an 
anatomical rather than morphological viewpoint include the wood, primary vascular 
system, laticifer systems, crystals, epidermis (including stomata and trichomes), 
floral vascular supply, sclerenchyma, pollen-grains and chromosomes. A further 
group of characters, such as colour, chromosomal behaviour during meiosis, antigen- 
antibody reactions and chromatographic analyses of various chemical constituents, 
in fact represent plant structure at a third (i.e. sub-microscopic) level : that of the 
molecule. Only a synthesis of all this and other evidence will provide anything 
approaching a complete picture of each species, and thus an ideal classification. 
Even today there is a great reluctance by many taxonomists to make real use of 
any but macroscopic characters, and there are a large number who believe that 
vegetative features are to be used only as subsidiaries to evidence from floral studies. 

Of all the types of somatic anatomy studied it seems clear that the less superficial 
layers of the plant have provided the best evidence for taxonomists, the wood, 
where present, perhaps being the most useful of all. The leaf epidermis, or its 
cuticle, has received relatively little attention from neobotanical taxonomists and 
most of the studies undertaken have been concerned with somewhat specialized 
groups of plants, e.g. Gramineae. Palaeobotanists, however, have used epidermal 
characters in classification since the beginning of this century, and today their use is 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 5 

practised as a routine procedure, although here the groups of plants concerned have 
usually been non-angiosperms. In recent years much more attention has been paid 
by both neobotanists and palaeobotanists to their use in angiosperms, but most of 
the studies are far from adequate and the subject is still in its infancy. 

Except in fresh material of a relatively small proportion of species the leaf 
epidermis of modern angiosperms is extremely difficult or impossible to remove 
in toto. In fossil species, moreover, it is frequently not or only fragmentarily 
preserved, and most of the early work on fossil leaf forms was carried out solely with 
the aid of leaf impressions. The discovery that the cuticle, which covers the whole 
of the leaf epidermis, bears an imprint of the epidermal features was therefore of 
great importance, especially as it often remains on fossil leaves in a well-preserved 
state long after all cellular layers have disintegrated. Moreover, except in plants 
with a very thin cuticular covering, the cuticle can be removed from the fossil or 
modern leaf (whether fresh or as a herbarium specimen) in a continuous sheet, to 
which end several methods have been employed. Whilst some epidermal features 
are lost in cuticular preparations the cuticle does provide some extra characters, 
and the study of the cuticle rather than the epidermis places the work on a consistent 
and precisely definable basis which is comparable with palaeobotanical investigations. 
The present survey is concerned with the cuticle itself, although epidermal studies 
are of course highly relevant. 

The level of knowledge now reached with regard to the systematic anatomy of 
the epidermis or cuticle is quite impressive in some groups, such as Gramineae or 
gymnosperms. In the dicotyledons and other monocotyledons, however, little 
extensive work has been attempted. Most investigations have been aimed at 
separating small groups of species or genera with little or no attempt at studying 
the degree of variation found within the taxa which are being separated. Much 
of the literature in the dicotyledons consists of short sections subsidiary to much 
wider studies of a more general nature. Work on fossil or recent plants by palaeo- 
botanists has been solely for the purpose of identifying fossils by means of a compari- 
son with living species rather than to investigate the intrinsic features of either 
group, and scarcely any mention is found of the possibilities of the presence of 
environmental or other variation. Many of the better surveys, even excluding 
the gymnosperms or Gramineae, appear to have been undertaken on plants with 
leathery, somewhat xeromorphic leaves (e.g. Lauraceae, Magnoliales). Such leaves 
have a thick cuticle which shows the epidermal cell pattern clearly, and for some 
reason these xeromorphs usually have a less variable pattern than mesomorphic 
species. Studies of the latter types have mostly been of a very limited scope, and 
in many cases a quite inadequate amount of material has been used as representative 
of each species. 

Thus there is a considerable gap in the literature with regard to both the variation 
of cuticular patterns within a. species and the types of characters and their degree 
of taxonomic usefulness which are to be found in " normal " mesomorphic di- 
cotyledons. The series of investigations upon which this article is based was aimed 



6 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

directly at helping to fill this gap. The survey was intended to be intensive rather 
than extensive, for, whilst this narrows the range of experimental material, it was 
just this close study of a limited group of plants which was needed. Extensive 
investigations have been undertaken in the past by a few workers, mainly palaeo- 
botanists, but they have invariably been too superficial for general use. The 
families of plants chosen for this programme, Comhretaceae, Rhizophoraceae and 
Avicenniaceae, are almost wholly tropical. They contain about 550, 120 and 11 
species in twenty, sixteen and one genera respectively, and although all are woody 
in habit the species show a great variety of growth forms, including trees, shrubs 
and lianes of all degrees of specialization occurring in all types of tropical habitat. 
All three families, of which the first two are regarded as very closely related but 
the third as very remotely so, contain mangroves. It is suggested that the results 
gained from the intensive study of these families have enabled a good number of 
general conclusions to be drawn which are applicable to all or most groups of 
dicotyledons. 

A soundly based theory of cuticular patterns can be put to a good number of 
uses besides the obvious applications in identification, taxonomic research and 
phylogenetic investigations. These include peat stratigraphy, pharmacognostical 
analyses and animal foodstuff research. Only after thorough examination of all 
aspects of epidermal and cuticular variation can systematic cuticular evidence be 
legitimately used in these fields, and it is hoped that the following sections will 
go some way towards realizing that aim. 

§2. Historical introduction 

A brief historical survey of the study of epidermal and cuticular anatomy throws 
much light on the situation at the present day, and is a considerable help in under- 
standing the areas in which work is now most needed. 

The earliest reference to fossil cuticles is apparently the note by Brodie (1842) 
" When the sandstone is freshly broken the epidermis of the fossil frequently peels 
off . . .", but this observation was not followed up. Three years later Goeppert & 
Berendt (1845) figured cuticular fragments of one species of conifer and two species 
of angiosperm which they found preserved in amber. These figures appear in a 
large volume primarily concerned, amongst plants, with twigs and leaf impressions, 
and this is the case with the later cuticular diagrams of Schleiden (1846), Unger 
(1853) and Wessel & Weber (1856). In all four cases the figures are crude and 
highly diagrammatic, and except where the material has been available for subse- 
quent study the true identity of the plants is obscure. For example the leaves 
described by Unger as a Potamogeton have been found to belong to a member of 
the Loranthaceae, whilst those figured by Schleiden are not identifiable at all 
(Edwards, 1935). Wessel & Weber described and figured a number of dicotyledons 
which they referred, in most instances, to living species, whilst Goeppert & Berendt 
more advisedly used generic names, such as Alnites, which were not applicable to 
living plants. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 7 

The first really important work on the systematic treatment of cuticles was that 
of Bornemann (1856), describing fossil cuticles of cycads from the neighbourhood 
of Thiiringen. He was the first person to realize the exact nature of the cuticles 
which he studied, and went to great pains to explain the relation between them and 
the leaves which bore them. He wrote : " The outline of the cell walls of the 
epidermis is almost always shown on the (cuticular) membrane by a network of 
dark brown lines, which apparently represent the cell walls, but which are much 
thinner than these would have been. These brown lines are to be regarded as 
parts of a homogeneous cuticle ..." (transl.). 

From then onwards a large number of works, from short notes to large volumes, 
began to appear on the subject of fossil cuticles. Even so, systematic studies on 
the surface features of angiosperms generally made use of the whole epidermis, 
and such investigations in the latter half of the nineteenth century were very 
numerous. One of the earliest of these was the investigation by Prillieux (1856) 
into the types of peltate trichomes found in various members of the Oleaceae, and 
he was actually able to separate all 23 species examined on the structure of these 
trichomes alone. 

The first taxonomist to make constant use of anatomical characters for his 
diagnoses was apparently Bureau (1864) in his revision of the Bignoniaceae, although 
these characters naturally formed but a very minor part of the descriptions. Soon 
after this anatomical characters began to appear more widely in taxonomic treat- 
ments, reaching a climax in Engler & Prantl's Die natiir lichen Pflanzenfamilien 
(1887-1915). Conversely anatomists were becoming more aware of the systematic 
value of many of their studies, and a number of extremely useful surveys appeared 
at that time, several of which are unknown to many present-day workers. Weiss 
(1865) produced a monumental account of the sizes, numbers and distribution of 
stomata in a large number of dicotyledons ; Bokorny (1882) surveyed the distri- 
bution and occurrence of pellucid spots in dicotyledon leaves, classifying them 
according to their causes ; Bachmann (1886) gave a beautifully illustrated account 
of all the types of peltate hairs known, in systematic arrangement ; Grob (i8g6) 
provided a very useful early account of grass-leaf epidermides ; and von Minden 
(1899) thoroughly covered the different types of water-secreting organs (e.g. water 
stomata, hydathodes) found in dicotyledons. It was perhaps this productive 
period which led Fritsch (1903) to summarize the most important systematic 
anatomical characters to which taxonomists might pay more attention, and in many 
instances his comments still apply today. Solereder's encyclopaedic volumes on 
dicotyledon systematic anatomy, first pubhshed in 1898-99, represent the epitome 
of anatomical works at that time, and together with a supplement were translated 
into English by Boodle & Fritsch (1908). They remain a standard reference today, 
one of their most valuable aspects being the detailed survey of the principal families 
in which a wide range of anatomical characters are to be found. 

Although works on fossil cuticles from Bornemann's (1856) time onwards were 
equally numerous very few of great value were produced before the present century, 



8 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

and scarcely any of these were concerned with angiosperms. Schenk (1869-71) 
included some drawings of cuticles in his studies of Wealden plants, but of course 
none of these was an angiosperm. Zeiller (1882) wrote what must be regarded as 
the first review on cuticular studies, and his article brings together much useful 
information with many supplementary original facts on cuticles of fossil ferns and 
conifers. Perhaps the most important of the early cuticular studies, however, were 
those of Nathorst (1907-12). His Paldohotanische Mitteilungen represents the first 
work which reported a long series of cuticular studies on a large number of species, 
and should be considered as the consolidation of this subject as a truly scientific 
discipline. 

One of the most important features of Nathorst's work is the improvement in 
various techniques, and these were quickly assimilated by palaeobotanists elsewhere 
in Europe and in America. This resulted in a considerable number of studies 
entirely devoted to cuticles, rather than studies using cuticular characters as sup- 
plementary evidence as had previously been usual. Thompson (191 2), for instance, 
studying Cretaceous conifers, re-examined the genus Frenelopsis and concluded that 
Zeiller (1882) was wrong in believing that its stomata had the unusual number of 
four guard-cells : in fact these four cells are the subsidiary cells, the guard-cells 
being quite normal. 

The first noteworthy cuticular investigations by British workers were those of 
Thomas & Bancroft (1913) on cycads. They compared cuticles of fossil and recent 
species with an emphasis on stomatal characters, and thus uncovered a considerable 
amount of detail regarding the relationships of the fossil types. Some of the first 
good photomicrographs of cuticles were included in this publication, and others 
were produced by Wills (1914) who reported on fossils from British coal-measures. 
Thomas (1930) later produced an article supplementing the knowledge of Mesozoic 
cycadean fronds set out in the earlier work. To exemplify the spread of cuticular 
studies from continental Europe Holden's (1915) paper may be cited, being one of 
the first works on Asian cuticles. From her study of Indian fossil conifers she 
concluded that epidermal leaf characters may often be of great value in specific 
determination, but that they seem of little use as phylogenetic evidence. 

By far the most important cuticular investigations are those of Florin, who for 
the past forty years has produced a large number of excellent accounts of recent and 
fossil gymnosperm cuticles, together with first-class photographs and drawings. 
One of his earliest important papers (Florin, 1920) was concerned with a comparison 
of fossil conifer cuticles with those of modern conifers, and most of his subsequent 
work has been developed along these lines. Florin's most important work on 
modern conifers (1931) includes descriptions of the cuticles of almost every known 
species, with particular reference to stomata, and a section is also included on the 
stoma types of the major gymnospermous fossil and recent groups for comparative 
purposes. Later (1938-45) he produced an even larger work on Upper Carboniferous 
and Lower Permian conifers, including full details of their cuticles, but this work 
also embraces the reproductive structures and other vegetative organs than the 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 9 

leaves. Perhaps his most important work on non-coniferous cuticles was the account 
of the stomata of Mesozoic cycads (Bornemann's subject in 1856) and Bennettitales, 
attempting to shed some light on the relationships of these groups (Florin, 1933). 
More recently Florin has devoted most of his time to studying conifer cone-structures, 
and it is fortunate that he continued to investigate conifer relationships rather than 
turn his attention to angiosperm cuticles as Edwards (1935) once suggested. 

Bandulska (1923) investigated the cuticular anatomy of dicotyledon and conifer 
leaves found in Eocene deposits at Bournemouth, and was able to place one of the 
latter into an already known fossil species. The dicotyledons, which were un- 
identified at the time, were placed into a new fossil form-genus, Dicotylophyllum. 
In her following four papers Bandulska described a series of cuticles of modern 
dicotyledons and in several genera she placed species which she had previously 
described under Dicotylophyllum. These genera were Nothofagus and Fagus in the 
Fagaceae (1924) ; Aniha, Lindera, Litsea and NeoUtsea in the Lauraceae (1926) ; 
Cinnamomum also in the Lauraceae (1928) ; and Tristania and Rhodomyrtus in the 
Myrtaceae (1931). This work was the first important investigation of modern 
dicotyledon cuticles to be attempted, and it clearly showed that with a great deal 
of patience conclusive and positive results can be obtained from this approach. 

Odell (1932) published a highly controversial paper which not only condemned 
the identification of fossil leaves by their gross form (shape and venation), but also 
severely attacked the use of cuticular characters for this purpose. She apparently 
originally attempted to identify the older Eocene fossils of the Bournemouth 
deposits, where Bandulska had first investigated the younger Eocene leaves, but 
concluded that the variation shown in all characters which were normally used to 
this end made it an impossible task. The bulk of her paper consists of a series of 
conclusions drawn from the study of 170 species of modem angiosperms, but un- 
fortunately she does not state the extent of her sample of each species. Her method 
of argument is to take each character used in cuticular differentiation separately, 
discuss the evidence for the fact that this character is known to vary under environ- 
mental or other conditions, or that completely unrelated species are identical (or, 
conversely, closely related species quite different) in that character, and conclude 
that it is taxonomically worthless. One could, of course, equally well do this with 
any other character that has ever been used taxonomically, but because (for example) 
Ranunculus ficaria L. possesses about seven to twelve petals one cannot conclude 
that petal number is of no diagnostic importance, either in this or in any other group. 

It is perhaps worth mentioning two of the twelve or so characters which Odell 
considered, as more or less random examples. With regard to the presence and 
types of subsidiary cells around the stomata she states that the three main types 
recognized by Solereder are not found especially in any taxonomic groups, but are 
distributed variously in different genera and families. Moreover in some species 
only the stomata on the stems are associated with subsidiary cells. Thus " the 
presence or absence of subsidiary cells bordering the stoma can be of little systematic 
value ". In the case of non-glandular hairs Odell mentions that the arrangements 



lo CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

and numbers per unit area on a leaf vary tremendously, and are affected by environ- 
mental factors such as sunlight, humidity, wind and altitude. Their abundance also 
varies on different parts of the same plant or even of the same leaf. Types of hair 
often differ within a family, and there is often more than one type even on one leaf, 
so that the types of hair are not connected with taxonomic categories. She therefore 
decided that " clothing hairs cannot be used in the identification of the vegetative 
parts of Angiosperms ", and that their " structure has been found of no diagnostic 
value ". Her final conclusions were that " any feature of the epidermis of the 
vegetative parts of living Angiosperms is unsatisfactory for diagnostic work ", and 
that " the modern method of naming fossil Angiosperms from a combination of the 
form, venation, and epidermal structure of their vegetative organs is quite inadequate 
for specific or even generic diagnosis". 

In almost all instances Odell's statement of fact must be regarded as correct, but 
it is in some of her extraordinary conclusions that she is obviously in error. If she 
was attempting to show that it is in general impossible to place a vegetative portion 
of a plant into its correct family or order on a single epidermal character (or even on 
a combination of such characters) then one would readily agree with her, but to 
assert that each of the twelve or so characters is "of no diagnostic value ", or 
" cannot be used in determinative work ", is obviously quite illogical. Odell 
appears to have overlooked two important basic points in drawing her conclusions. 
First of these is that the compilation of diagnoses and hence the identification of 
species should not be undertaken with a single character, but with a combination 
of as many characters as possible. Although two families, genera or species may 
not be separable by a single character a combination of several characters will often 
enable separation to be effected. Secondly a range of form is often more important 
taxonomically than is a fixed type. For example the possession of a variable number 
of petals by Ranunculus ficaria is a notable taxonomic feature of this species, not a 
reason to doubt the systematic value of petal number. Further criticism of Odell's 
remarks is unnecessary. In general one would agree with her that cuticular characters 
are usually not useful at the family level or above, but it is unfortunate that she does 
not differentiate between those cases where the characters are useful in phylogenetic 
classification and those where they are useful for identification. 

Belying Odell's claim that stomatal structure " cannot be regarded as of systematic 
value " just four of several publications may be cited. Carolin (1954), working 
with Dianthus, found that contrary to the usual findings (see below) neither stomatal 
size nor frequency could be used to indicate the level of ploidy. However, within 
the diploid and within the hexaploid groups most investigated species could be 
separated by a series of size, frequency and morphological characters. Stebbins & 
Khush (1961) surveyed the " stomatal complex " of most families of monocotyledons. 
They found that characters such as the number and type of subsidiary cells gave 
valuable clues to the affinities of the plants, so that each family had a predominant 
type or arrangement. Watson (1962) found that in the Epacridaceae the stomatal 
distribution and orientation on the perianth and leaves, and also the broad structural 
types of stomata, were mostly differentiated at the generic or a higher level, and in 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY ii 

the Ericaceae "(Watson, 1965) the stomata provided equally important taxonomic 
evidence although in several instances there was little or no correlation with the 
usually accepted classification. 

Pharmacognosists have frequent need to identify leaves or fragments of leaves 
used as herbs, and, as the number of species is relatively small, epidermal features 
have often made identification possible. Timmerman (1927) produced an early 
analysis of one of the characters used in pharmacognosy, namely stomatal frequency. 
Using Datura she found that stomata were more frequent at the leaf apex than in 
the middle, whilst the ratio on the two epidermides also altered in different parts of 
the leaf. She therefore concluded that stomatal frequency was generally of no use for 
identifying leaf fragments in this genus, although she was able to separate one species 
from the rest of those examined by this character. In the same year, working from 
a different angle, Salisbury (1927) also investigated the constancy of stomatal 
frequency, adding environmental effects to the positional factors considered by 
Timmerman. Salisbury found much the same variation as Timmerman, except that 
the proportions on the two epidermides did not appear to alter. He found that if 
the proportion of stomata to epidermal cells (the stomatal index, i.e. lOoS/E + S) 
was measured instead of the stomatal frequency the variation produced by environ- 
ment and position could be completely cancelled, apart from the effect of humidity. 
He was able, in fact, to measure the environmental humidity by means of the 
stomatal index. This very simple information was of extreme importance to 
taxonomists, and the stomatal index is now used as a routine procedure in palaeo- 
botany (cf. Harris, 1944). Salisbury's use of the stomatal index had been fore- 
shadowed, however, by Loftfield (1921), who noticed that the ratio of stomata to 
epidermal cells often remained constant when the stomatal frequency varied. 

A less reliable method of overcoming the variation in stomatal frequency was also 
used by Baranov (1924), who found that the stomatal frequency of the middle 
portion of leaves half-way up the stem was equal to the average frequency of all 
parts of all leaves. This does not take into account the fact that most fossils are 
only fragmentary, nor the effects of environment. 

Working much later than the above, Gupta (1961) also found a considerable 
variation in stomatal frequency on leaves from different parts of the plant, and 
demonstrated a negative correlation between this and the lamina size. He thus 
used an Absolute Stomatal Number, the product of the stomatal frequency and the 
lamina size, as a constant, this being comparable to the stomatal index. 

Sax & Sax (1937) and Sax (1938) reported on some very interesting and now almost 
classic studies on the effect of polyploidy on stomatal size and frequency. They 
found a positive correlation between each of these two variables and polyploidy, and 
could usually separate the diploids and tetraploids in any given series of plants. 
Modern Floras now often contain characters such as stomatal size and stomatal index 
in the separation of closely related species where one is a polyploid (e.g. Rorippa 
nasturtium-aquaticum (L.) Hayek and R. microphylla (Boenn.) Hyland.). 



12 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

The great mass of anatomical work which marked the end of the last century 
seems to have largely ceased with the publication of Solereder's work and supplement 
(igo8). Rehfous (1917), however, provided a very useful survey of work on 
stomata. In the 1930's systematic anatomy, together, in fact, with taxonomic 
studies in general, began to regain popularity, and from then on the number of 
systematic studies has increased quite steadily. 

Meyer (1932, 1932a) investigated the Alismataceae, firstly for taxonomic and se- 
condly for phylogenetic reasons. Most of the genera and species Meyer examined 
could be separated by anatomical features, but he found little evidence suggesting 
close relation between the Alismataceae and Ranunculaceae ; the stomata, for 
instance, are quite different. This work was connected with the commencement of 
a Systematische Anatomie der Monokotyledonen by Solereder & Meyer (1928-33), of 
which only a very small proportion appeared. 

Prat (1932) published an extremely important work surveying the systematic value 
of grass epidermides, this family without doubt being the most amenable amongst 
the angiosperms to epidermal classification. Several important taxonomic changes 
have been foreshadowed in the Gramineae by epidermal considerations, such as the 
exclusion of Nardus from the Hordeeae or of Eragrostis from the Festuceae, a situation 
only elsewhere paralleled in some gymnosperms. Many more recent works on this 
subject have appeared, but the exceptional nature of the Gramineae precludes the 
necessity of their being considered fully here. The most important work is that of 
Metcalfe (i960), which is a complete survey of almost all genera and of all the litera- 
ture. Some recent papers which show the amazing systematic value that can 
confidently be placed on minute features of grass epidermides are those of Tateoka 
et al. (1959) and Borrill (1961). In all cases, however, students of grass epidermal 
anatomy have used the whole epidermis in preference to the cuticle. 

In the last twenty years a great many systematic anatomical studies have been 
undertaken, and the resulting publications provide sections on epidermal and 
cuticular characters varying in length with their taxonomic importance in the group 
concerned. Mention may be made of the work of Bailey & Nast (1944, 1948) in the 
Winteraceae, Illiciaceae and Schisandraceae, Heintzelman & Howard (1948) in the 
Icacinaceae, Morley (1953) in the Melastomataceae, Tomlinson (1959) in the Musaceae, 
and Paterson (1961) in the Epacridaceae. The degree of use which can be obtained 
from the epidermis varies considerably in these groups, as would be expected. Thus 
Tomlinson was able to use a number of general and special epidermal characters to 
great effect in delimiting at the generic level, but Bailey & Nast, who were two of 
the few anatomists to isolate the cuticle and study it as such, found that apart from 
its very distinctive appearance throughout the Winteraceae it did not furnish good 
taxonomic characters. 

Over forty years after Solereder's volumes (1908) Metcalfe & Chalk (1950) published 
along similar lines a modem survey of dicotyledon systematic anatomy. The 
hterature survey is essentially concerned with work done since Solereder's time, and 
many details of structure mentioned by the latter are omitted where no further 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 13 

examinations have been undertaken. The greatest improvements and additions are 
to be found in the sections on secondary-xylem structure, upon which emphasis is 
justifiably placed. With regard to the epidermis and cuticle Solereder's work is 
therefore frequently the more informative. In reviewing characters of taxonomic 
importance Metcalfe & Chalk obviously consider wood and associated tissues to be 
of the highest importance, and epidermal characters are dismissed in a few paragraphs. 
With regard to stomata they quite correctly state that the number per unit area is 
too variable to be of any great importance, but it is strange that they do not mention 
the usefulness of the stomatal index. More recently Metcalfe has commenced a 
comparative Anatomy of the Monocotyledons, of which to this date volumes have 
appeared on the Gramineae (Metcalfe, i960) and the Palmae (Tomlinson, 1961). 
Without such an important basis as Solereder's dicotyledon volumes the compilation 
of a systematic anatomy of the monocotyledons is a considerably greater task, and 
when completed will probably prove even more valuable. 

In 1954 Metcalfe published a paper which may be compared with one by Fritsch 
(1903) half a century before. This set out to indicate the main lines along which 
systematic anatomy could be developed, and in the next year (1955) he summarized 
recent work on this subject in the monocotyledons, especially the grasses. More 
recently Metcalfe (1963) has elaborated his opinions on the desirable trends in 
modern comparative plant anatomy. 

From about the time of Bandulska's work (1923-31) onwards a considerable 
number of articles on fossil dicotyledons began to appear in the literature, in several 
cases obviously directly inspired by her publications. Stockmans (1932) reported 
on two species discovered in Belgium : a Litsea and a Dewalquea. The first genus, 
a member of the Lauraceae, had been discovered in England by Bandulska (1926), 
whilst the extinct genus Dewalquea had been previously found in Ireland by Johnson 
& Gilmore (1921). Discussions by Stockmans and Johnson & Gilmore, in both cases 
well illustrated, on the affinities of this genus produced little concrete theory, several 
diverse families being mentioned. Since not all of the species of Dewalquea possessed 
peltate scales on the leaf it was suggested that they were perhaps representatives 
of two different genera. Whether or not this is true it must be remembered that other 
genera known today (e.g. Combretum and Rhododendron) possess scaly and scale-less 
species. 

Hofmann produced a valuable series of papers from 1926 to 1932 (cf. Hofmann, 
1932) on plant remains in early Tertiary formations. In her earlier works she 
attempted to give names of living genera or species to the cuticles, and in 1926 she 
described fossil plants in ten living genera from these deposits. In 1932, however, 
she changed to the more advisable system of describing these old leaves in a fossil 
form-genus. Her choice of name for this was Folium, which is nomenclaturally 
superfluous owing to Bandulska's early name Dicotyhphyllum for the same purpose. 
Edwards (1935) justifiably comments that her descriptions of new fossil leaves are 
somewhat inadequate, but she gives some good photographs. 

Kubart (1927) found great difficulty in distinguishing between the fossil leaves of 



14 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

Fraxinus (Oleaceae) and Umbellaria {Lauraceae) which he discovered in North 
America. He pointed out, however, that had the cuticles been preserved there 
would have been no difficulty in discrimination as the epidermal structure of the 
two genera is quite different. In support of this he figured a series of recent species 
of each genus, showing the great divergence between them. Also working on Ter- 
tiary plants, Straus (1930) was able to identify a number of species with modern 
genera, and in the cases of the two better-preserved genera [Juglans and Populus) 
he gave some good high-powered photographs. Straus pointed out the need for 
microscopic examination of the cuticle in cases where two fossil leaves appeared to 
be closely similar. 

In recent years the great majority of works on fossil cuticles has been concerned 
with pre-Tertiary plants, mostly gymnosperms. Perhaps the most important 
publications have been those of Harris (i942->-), who has covered a wide range of 
topics on the Jurassic flora of Yorkshire in several scores of articles, many of these 
including cuticular studies which have often proved of considerable diagnostic value. 

Cookson (1953) reported on a relatively recent (early Tertiary) cycad from Australia 
which, on the basis of its cuticular characters, could be referred to the genus 
Macrozamia, which still occurs in Australia today. Cuticles of this fossil and of the 
living M. hopei W. Hill ex F. M. Bailey are shown, and due to the great similarity 
between them the former is named M. hopeites Cookson. 

Apart from Europe and North America some of the most important cuticular 
studies have emanated from India, and as examples two of the papers on the Indian 
Glossopteris flora by Srivastava and his colleagues may be cited. Srivastava (1956) 
produced a most thorough account of all the species of Glossopteris, Gangomopteris 
and Palaeovittaria, with very clear descriptions, drawings and photographs. All 
species could quite easily be separated on cuticular characters alone, a whole series 
of features, concerning epidermal cells, stomata and venation, being employed. 
Surange & Srivastava (1956) concurrently considered the generic limits in this group 
of leaves. They concluded that the three genera, which were delimited on frond 
shape, could be better divided into six genera using the cuticular characters. These 
six genera did not correspond in any way to the three established genera, and 
pending further information no new generic names were proposed. 

Apparently the only general review of cuticular studies in angiosperms is that of 
Edwards (1935). He gives an extremely useful survey of the literature to that date, 
which has obviated the necessity of very detailed accounts of the earlier work in 
this introduction. The account treats recent and fossil monocotyledons and dicoty- 
ledons in four separate sections, those on monocotyledons being mostly concerned 
with grasses. Many of his citations are annotated with some sort of assessment on 
their importance or validity, but in the case of Odell's (1932) paper he quite justifi- 
ably records strong protest with a series of objections, mostly in the form of counter- 
evidence. One would concur with almost all of his opinions on this and other papers, 
but as mentioned previously I prefer to attack Odell's deductions and conclusions 
rather than her statement of fact. In several instances it is obvious that Edwards 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 15 

held almost as extreme ideas as Odell, though in a completely opposite direction. 
He was clearly convinced of the tremendous and widespread usefulness of cuticular 
characters in the angiosperms, which is perhaps to be expected of a palaeobotanist 
whose cuticular studies had mostly been concerned with gymnosperms. Had he 
made extensive investigations in the angiosperms he would have met, in many 
instances, with a much wider range of variation than is found in any gymnosperm, 
and than he seemed to admit as existing in the angiosperms. For instance, in 
criticizing Odell's statement that the leaves of Campanula latiloba A. DC. and Inula 
salicina var. denticulata Borbas are identical, he found that there were in fact good 
microscopic differences between them, and he also suggested that Odell's " own 
drawings do not suggest identity of epidermal structure ". Odell's figures differ 
only in the thickness and degree of undulation of the epidermal cell walls, and had 
Edwards been familiar with the range of variation of the cuticles of many angio- 
sperms he would have realized that most species show a wider range of variation 
than Odell's two drawings. Thus, although in fact the two species do apparently 
differ in cuticular structure {fide Edwards), Odell's figures certainly do not suggest it. 
Apart from a few criticisms such as the above, Edwards's paper is a valuable and 
critical review. His conclusion may be quoted : " given careful and critical work 
on well-preserved material, together with a detailed comparison with a wide range 
of living forms, results obtained from a study of fossil angiosperm cuticles will be as 
valuable as those derived from any other fossil remains, and certainly far more 
reliable than those founded on leaf impressions alone ". 

Edwards did not mention any articles concerned solely with modern dicotyledon 
cuticles, and the first that the writer has traced is the account by Rao (1939) of the 
order Magnoliales in the sense of Hutchinson (1926), a very heterogeneous assembl- 
age of families about which Rao concludes: " From the point of view of epidermal 
studies, the Magnoliales do not represent a co-sanguinary group, but indicate a 
convergence ". Rao's paper presents a detailed description of the cuticles of over 
50 species representing more than half the genera of the order, but since he admitted 
that in most cases only a single leaf of each species was examined the results must 
be treated with extreme caution. Objection must be made to some of Rao's con- 
clusions, although a detailed criticism will not be attempted here. Rao implied, 
for instance, that the cuticular striations found in some species of Magnoliales are 
phylogenetically related to those of the Cycadales. However, representatives of 
many dicotyledon families possess cuticular striations which are identical with those 
of the Magnoliales. Even if cuticular striations in the two orders were phylogeneti- 
cally related, and this seems unlikely and in any case could never be proved, no such 
inference could be drawn merely from their presence in both groups. Rao strongly 
stressed further evidence of a connexion between the Magnoliales and gymnosperms 
from his stomatal studies. Using Florin's (1931) two categories of stomatal de- 
velopment Rao claimed that two genera of Magnoliales, Euptelea and Cercidiphyllum, 
possessed haplocheile stomata, which according to Rao were otherwise virtually 
unknown in the dicotyledons yet were typical of all gymnosperms except the 
Bennettitales, Welwitschiales and Gnetales. Later work has shown many of Rao's 



i6 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

conclusions regarding stomatal types to be erroneous, and a modern stomatal survey 
of the Magnoliales is badly needed. Jalan (1962), for example, has found that the 
stomata of Schisandra are haplocheilic (as, in fact, they are in a great many 
dicotyledons), and the same is true of Gnetum (Maheshwari & Vasil, 1961). In 
Magnolia and Michelia species the stomata on the leaves are syndetocheilic, but 
those on the flowers haplocheilic (Paliwal & Bhandari, 1962). The need for develop- 
mental studies rather than observations of mature structure cannot be over- 
emphasized, and the approach of several Indian workers in this field is encouraging 
(cf. Pant, 1965). Further reference to this is made in the section on stomata. 

Ahmad (1964, 1964b, 1964c) has investigated the cuticular and epidermal anatomy 
of 17 genera of Solanaceae, and found that a number of characters concerning the 
stomata, trichomes, epidermal anticlinal cell walls and cuticular striations are very 
useful in identification. The same characters enabled 28 species of Solanum and 
23 of Cestrum to be differentiated. 

Sinclair & Dunn (1961), on the other hand, working in the monocotyledons, 
appear to have made conclusions regarding the usefulness of cuticular characters 
at the family and generic levels which are not justified on present information. 

It is perhaps largely by chance that the Lauraceae have been studied more 
frequently than any other dicotyledon family with regard to their cuticular charac- 
ters. Studies already mentioned have been those of Bandulska (1926, 1928), Kubart 
(1927) and Stockmans (1932), and there have been other less important ones besides. 
Kostermans (1957), in his taxonomic studies of the Lauraceae, mentioned the useful- 
ness of cuticular characters, but pointed out the care necessary in the interpretation 
of the results of these investigations. He was referring mostly to Bandulska's work. 
Marlier-Spirlet (1945) investigated the epidermal anatomy of Cinnamomum more 
thoroughly than did Bandulska, and produced many figures. He also showed 
vertical sections of the leaf to demonstrate the derivation of various features of the 
cuticular preparations which were not readily interpretable from surface views 
alone. It was found to be possible to separate many of the eighteen species in- 
vestigated, and three broad groups were recognized, containing four, three and 
eleven species respectively. The most useful taxonomic criteria appear to be the 
shape of the upper epidermal cells and the types of stomata. A continuation of 
this study would be of considerable interest and value. Dilcher (1963) undertook 
a " cuticular analysis " of an Eocene species of Ocotea, and after a study of its 
variation and of the living taxa concluded that it should be considered a distinct 
species. 

Martin (1955) used cuticular anatomy to a most novel and useful end to which 
there appear to be no subsequent references. Previously Parkinson & Fielding 
(1930 ; fide Martin, 1955) had compiled notes of some epidermal characters of the 
plants used as cattle food, using the solid leaf fragments. In order to study the 
diet that sheep choose in various pastures Martin examined the rumen contents 
and faeces of these animals to ascertain quantitatively the species of plants re- 
presented by their cuticles. In the upland pastures in which the work was pursued 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 17 

the number of species was rather Hmited and few species were closely related ; thus 
Martin was able to separate all forty likely plants by cuticular anatomy and deter- 
mine them quantitatively in the faeces. He gives brief descriptions and photographs 
of twelve representative species. Of these forty species only sixteen were recognized 
in the stomach contents or faeces, including four grasses, two sedges, three rushes, 
two mosses and five dicotyledons. Except in the case of two species of J uncus 
each genus was represented by only a single species, although other genera were 
represented by two or more species among the originally investigated group of 
forty plants. It seems likely, however, that some or many mesomorphic herbs 
would never be represented due to their extremely thin cuticle, although the only 
known cutinase enzyme has been isolated for certain solely from PenicilUum 
spinulosum Thom (Heinen, i960, 1961), 

This work is a further indication of the economic use to which cuticular studies 
may be put, and emphasizes the need to discover the basic principles underlying 
the possible application of this method in various directions. 

§3. Anatomical introduction 

Some idea of the precise relationship between the cuticle and the underlying 
epidermal cells is essential to a full understanding of the nature of the patterns seen 
on acellular cuticular preparations and their interpretation in terms of a cellular 
layer. 

Although Bornemann (1856) obviously understood the cuticular-epidermal 
relationship fully the first good general anatomical accounts are those of de Bary 
(1871, 1884), and Hohnel (1878) appears to be the first to have described the 
chemistry of the cuticle. During the past thirty years our knowledge of the cuticle 
and of its essential constituent, cutin, has steadily grown, accelerated by two major 
technical advances. The first of these, polarizing microscopy, became available in 
the 1920's and its use is shown in the work of several Continental writers. The 
second, electron microscopy, a post-war development, has been utilized in cuticular 
studies mainly by Roelofsen and a number of American workers. 

In all vascular plants the basic chemical component of all the cell walls is cellulose, 
whose long-chain molecules exist as long compound microfibrils. In the inter- 
fibrillar spaces are a number of substances, including other polysaccharides, fatty 
materials, water and inorganic and organic solutes. Separating two such walls of 
adjacent cells is a cellulose-free layer known as the middle lamella, consisting 
basically of calcium and magnesium pectates. The outer face of an epidermal cell, 
however, is unusual in being adjacent to the environment rather than to another cell, 
and it is on this face that the cuticle is found. The precise structure of the latter 
was first ascertained by means of polarizing microscopy (e.g. Meyer, 1938 ; Roelof- 
sen, 1952) and was confirmed later by electron microscopy (Roelofsen, 1959). 

External to the normal cellulose cell wall of the outer face of each epidermal cell 
is found a thin layer of pectic material which is presumably continuous with the 
middle lamellae of the anticlinal walls of the underlying epidermal cells (Fig. i). 



i8 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



Outside this is a usually two-layered wall, which according to the nomenclature of 
Roelofsen (1952, 1959) may be termed the cuticular membrane. The inner portion 
of this, the cuticular layer, is composed essentially of a cellulosic framework between 
the microfibrils of which are encrusted large amounts of cutin. The outer part is a 
usually thinner layer, the cuticle proper, which lacks cellulose and is composed 
mainly of cutin, which is adcrusted on to the cuticular layer. Esau (1953) refers to 
the process of encrustation as cutinization, and that of adcrustation as cuticulari- 
zation. The above terminology is adopted hereinafter. In a number of plants, 
usually those with leathery leaves, the four layers exterior to each epidermal cell 
lumen are simplified insofar as a cellulose-free pectic layer is absent, the cutin-free 
innermost cellulose wall gradually merging into the outermost cellulose-free cutin 
layer. Rarely are all four layers as sharply defined as suggested in Fig. i, but in 
general it can be said that the innermost layers lack cutin and the outermost layers 
lack cellulose. Chemically cutin consists of a number of highly polymerized long- 
chain hydroxy-fatty acids. 



cuticle 



cutin only" 



cuticular 

layer 
pectic laye 

cellulose 
cell wall 






cellulose 



>! increasing ;" 



;;L, "!''i'??^...R>^$;'i^^vvv>^ cellulose ^ 
? cellulose ''^f-^'il^i^ij^^ti:.-''^^ 



;-;increasing 

2 cutin 

'■Cy/^.i-:\ 

■.'j::;~:,-: 




cell outline 

observed on 

cuticular preparations 



extent of origina 
cell wall 



middle lamella 



original extent 
of cell walls 



Fig. I. The organization and interpretation of cuticular membranes and preparations : 
sections of epidermides with normal and abnormally extensive cutinization are shown, 
together with diagrams showing the significance of the patterns observed on the 
respective cuticular preparations. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 19 

Where the outer walls of two epidermal cells meet, the pectic layer, if present, 
may form a T-shaped juncture with the middle lamella of the vertical walls of the 
cells. In most cases, however, the cuticular layer projects down between the two 
cellulose cell walls for some way, forming the three-dimensional network of so-called 
cuticular flanges (Fig. i). The length of the cuticular flanges varies from their 
virtual absence to their reaching all the way to the inner wall of the epidermis. In 
some leaves a greater proportion of the epidermis may be cutinized, in extreme 
cases the cell walls on all faces of the epidermal cells and even the outer walls of 
the subepidermal layer being encrusted with a certain amount of cutin. This 
internal cutinization is a feature of many xeromorphic plants. 

Cuticular flanges are formed by the cuticular layer alone, and the cuticle remains a 
thin uninterrupted membrane. Indeed, the cuticular membrane as a whole is a 
continuous and virtually extra-cellular sheet, and should be thought of as something 
additional to the cell wall. It completely covers the leaf, usually including any 
trichomes which may be present, being interrupted only by the stomata. In a 
herbaceous plant it appears to surround the whole of the aerial shoot, including 
the apical meristems and flowers, although some workers have stated that certain 
trichomes lack a cuticular covering. With regard to the root system there is less 
certainty, opinions differing whether it is only the root tip and root hairs or nearly 
all the root that lacks a cuticular membrane (Lee & Priestley, 1924 ; Priestley, 
1943 ; Esau, 1953). At each stoma the cuticular membrane follows the contours of 
the guard-cells, and often ends on their inner walls. In some cases, however, it is 
more extensive, covering also the adjacent mesophyll cells, Cutinized cell walls 
have also been found in more internal regions of the plant (cf. Arzt, 1934), and 
Lamarliere (1906) found a " cuticule interne " lining the aerenchyma of water plants. 

Right from the time that the leaf is a minute primordium in the apical meristem 
it is covered by the rudiments of the cuticular membrane in the form of a very thin 
and pliable yet continuous fatty layer. These fatty substances obviously migrate 
through the cell wall from the cytoplasm, although there are no visible pores in the 
former for the passage of the cutin. On exposure to air the cutin hardens to a 
varnish-like cuticle by polymerization. A reasonably hardened cuticular membrane 
is present long before the epidermal cells have attained full size, and the problem is 
thus one of increasing the surface area of a relatively non-stretchable substance. 
Some workers suggest that the epidermal cells may grow in surface area only at 
their edges, so that the already hardened cuticular membrane remains undisturbed 
(e.g. Schieferstein & Loomis, 1959). 

Not only is wax present within cell walls and cuticular membranes but it is also 
found in a fairly pure state on the surface of the cuticle of many plants. In recent 
years these deposits, which are the cause of the " bloom " on the surfaces of many 
leaves, have been studied in considerable detail, since they are important to the 
understanding of water repellency. Fairly comprehensive accounts are given by 
Mueller et al. (1954), Schieferstein & Loomis (1956, 1959), Scott et al. (1958), Juniper 
& Bradley (1958), Roelofsen (1959) and Juniper (1959, 1959a, i960). It is 



20 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

interesting to note that the type of deposit (frequently variously shaped rods and 
plates) appears to be largely characteristic of each plant species. The deposits, 
when washed, dissolved or brushed off growing leaves, re-form quite rapidly and in 
exactly the same form as previously, and although their abundance varies with the 
environment and other factors their non-varying form is obviously in some way 
genetically regulated. What determines the form which the wax assumes is of 
course unknown, and moreover different parts of the plant (e.g. stem and leaf 
epidermides) may have totally different patterns. This subject is obviously of some 
taxonomic interest though doubtfully of any widespread practicability, and it is not 
considered further in this review. 

In studying the cuticular membrane various methods may be employed to isolate 
it from the rest of the leaf. All make use of one or both of two important structural 
features of the epidermis which will have become apparent from the preceding 
paragraphs : firstly, the cuticular membrane is in most plants separated from the 
cellulose cell wall within by a thin layer of pectic material ; and secondly, the 
cuticle and the bulk of the mature cuticular layer are essentially composed of a 
substance (cutin) of very different chemical structure from the main constituent of 
the cell wall (cellulose). Isolation of the cuticular membrane may be effected with 
either of two types of substance : an enzymic preparation, either a pectinase which 
digests the pectic layers and so frees the cuticular membrane from the underlying 
cells and these from each other (Orgell, 1955), or a cruder hemicellulase-cellulase- 
pectinase mixture which additionally digests the cell walls (Skoss, 1955) ; or a 
macerating solution, composed usually of strong acids, which completely dissolves 
away all polysaccharide leaving only cutin unaffected. Very much later the acid 
also dissolves the cutin. The enormous resistance of cutin to the effects of age, 
micro-organisms and chemicals is, in fact, its greatest attribute with regard to the 
present study. Only one organism {Penicillium spinulosum) is known to be able to 
break down cutin (Heinen, i960, 1961). The advantage of the above enzymic method 
is that the cutin is always completely unchanged, and in biochemical studies of the 
cuticular membrane this method is the only one to be recommended. On the other 
hand the macerating solutions are very much quicker and simpler to use, and the 
speed of reaction may be adjusted more easily. The cuticular membranes are also 
better for microscopic examination since they are considerably cleaner. 

The upper surface of such preparations is smooth, or minutely striate or papillose, 
whereas the lower surface is protracted into a reticulate series of flanges which 
represent the position of the epidermal cell walls. When viewed from above or 
below the cuticular membrane is much thicker at the position of these flanges, which 
therefore show up, especially if the cutin is stained, as a series of narrow darker 
lines representing the shapes of the original epidermal ceUs (Fig. i). It is important 
to realize that these lines only show the position of the central region of each double 
cell wall, and that the original walls were present for a considerable thickness on 
either side of them. In those cases where there are no cuticular flanges the cuticular 
preparations will of course lack the cell outlines, but it is apparently rare in woody 
plants for at least the faintest and most fragmentary of outUnes not to be present. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 21 

As mentioned previously some plants have a heavily cutinized epidermis, the whole 
of the outer, vertical and inner walls and even the top of the vertical walls of the 
subepidermal layers being encrusted with cutin. In these cases the preparations 
have cuticular flanges which are very long and join up at their bases, and in surface 
view a second fainter outline, that of the subepidermal layer, may be visible below 
the epidermal outline. There is some evidence that the cuticular membrane is more 
or less completely lacking in cell outlines, or is not even able to be isolated in a 
continuous sheet, in many though by no means all herbaceous dicotyledons. In 
these cases the whole epidermis, or plastic moulds of it, must be studied. 

§4. Materials and methods 

For the purposes of obtaining cuticular preparations leaves may be used in any 
state (fresh, spirit-preserved or dried) and it is fortunate that herbarium specimens 
even over 100 years old produce at least as satisfactory preparations of the cuticular 
membrane as fresh leaves, exactly the same techniques being used. It is important, 
if studies of cuticular membranes are to be used as routine taxonomic procedure, that 
the methods involved in their isolation should be simple, quick and reliable. 

With regard to the choice of material it is essential to ensure that a representative 
sample of each species is examined, and this involves a considerable amount of 
preliminary investigation into the variability of the taxa concerned. Thus the 
amount of material which can be considered truly representative will vary greatly, 
but in the present study I found it necessary to examine only one leaf of each 
specimen, except in cases of obviously different-aged leaves. The number of 
specimens of each species which was investigated varied according to the amount of 
material available, the degree of macroscopic variation apparent, and the ecological 
and geographical range of the species concerned. Three to five specimens were 
usually found to be a sufficient sample, but care was taken to select those with the 
widest morphological and ecological variation. In many cases, however, fewer 
specimens were available, while on the other hand special circumstances sometimes 
necessitated the examination of up to twenty. 

The first task in investigating the taxonomic use of a new character is obviously 
to check it with the other characters already used, and for this correctly named 
material is absolutely essential. In the present series of investigations the families 
concerned are rather well covered in the literature, although the lack of knowledge 
concerning two genera necessitated a complete revision of all the species (Exell & 
Stace, 1963), and in some families this problem is so marked that it would become the 
limiting factor. Other genera of the three families studied still badly require 
revision. Whatever the state of knowledge of the classification of the taxon 
concerned differences of opinion may exist and occur, and a careful record of the 
actual specimens examined is essential. All the permanent slides prepared by 
me are marked with the collector, the number, year and country of collection, and 
the herbarium from which the material was obtained. This enables the identifica- 
tion to be checked whenever required. 



22 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

Many different macerating techniques have been employed to isolate the cuticular 
membrane. Enzymic maceration is not suitable for taxonomic work as the condi- 
tions (temperature, pH, concentration) have to be carefully controlled and the method 
is very slow, and in fact in many thick leaves a clean preparation cannot be obtained. 
Of the milder chemical macerating solutions Eau de Javelle (CaCla and KgCOg 
solutions) is the most used (Nathorst, 1907 ; Bather, 1908 ; Bandulska, 1923), 
but for present purposes strong acids are preferable. Schulze's Solution (Schulze, 
1855), concentrated HNO3 with a few KCIO3 crystals added, is the most popular, 
and several improvements and variations have been suggested (e.g. Bather, 
1908 ; Thomas, 1912 ; Thomas & Bancroft, 1913 ; Gothan, 1915 ; Harris, 1926). 
I found, however, that a slight modification of Jeffrey's Solution (Jeffrey, 1917), 
equal parts of 10% chromic and bench concentrated nitric acids, gave the best 
preparations in the shortest time, and it has been used without exception. This 
mixture was variously diluted to conveniently regulate the time taken for macera- 
tion to about sixteen hours, i.e. overnight, and in all 250 or so species examined the 
method was entirely successful. Care must be taken not to allow maceration to 
continue for too long since over-macerated cuticular membranes are unsuitable for 
future manipulation and examination. 



Parts of leaf macerated 



Cuticular preparations 
obtained 



1 1 


r — 1 


1 " 

1 
1 


1 1 
-1 




margin 



midrib 

u = cuticular membrane from upper epidermis 
/ = cuticular membrane from lower epidermis 



Fig. 2. Diagram illustrating the origin of cuticular preparations. 



For purposes of conformity identical regions of each leaf were macerated (Fig. 2), 
so that three pieces of cuticular membrane were made into separate permanent 
mounts. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 23 

In some cases it is necessary to examine the cuticular pattern of species not avail- 
able for maceration, or of species whose cuticular membrane cannot be removed as 
a continuous sheet. For this purpose the preparation of a plastic epidermal cell 
mould can prove of value, the plastic, usually cellulose acetate or nitrate, being 
either painted on to the leaf as a solution or pressed on as a thin sheet and then 
partly and temporarily redissolved (e.g. Nathorst, 1907, 1907a ; Bather, 1907 ; 
Thomas, 1912 ; Long & Clements, 1934 ; Sinclair & Dunn, 1961), making use of 
the fact that the outer surface of the cuticle usually follows the contour of the 
epidermal cells to some extent. The version of this method considered most 
suitable is that of North (1956), who used the formula 2 gms. of cellulose acetate 
and 0-7 gms. of gentian violet in 100 mis. of acetone, the stain improving the contrast. 
Herbarium specimens, whether dry or boiled, rarely give such good results as fresh 
leaves, but they are usually quite usable. The method is useless for most thick- 
cuticled leaves, which do not show the lines of the epidermal cells on the outer 
cuticular surface, and for pubescent leaves, although in the latter case the hairs 
can sometimes be removed. Occasionally transparent cell moulds are found already 
prepared on herbarium specimens in the form of flakes of old dried herbarium gum, 
which may show the epidermal cell outline very clearly. 

Frequently sections of leaves must be prepared in order to interpret the observed 
cuticular patterns. Fresh or spirit-preserved leaves are naturally easier to section, 
but herbarium leaves of great antiquity, after boiling in water for a while, usually 
provide good sections when cut by hand or with a freezing-microtome. It is often 
more expedient to isolate the whole epidermis rather than the cuticular membrane 
alone, this method being particularly applicable to Gramineae. Lactic acid is 
generally used to separate the epidermis from the subepidermal layers, and the 
epidermis may then be stained with any general botanical stain (Clarke, i960). 

The technique employed to prepare permanent mounts was again designed to 
meet the three requisites of speed, simplicity and reliability. The best stain was 
found to be Sudan IV (1% solution in 70% alcohol) which has the advantage of 
being progressive, so that overstaining does not occur, but the disadvantage that 
dehydration and thus mounting in canada-balsam or similar media is impossible 
since the stain is re-dissolved from stained preparations by absolute alcohol. Thus 
glycerin-jelly was used as the mountant, and although this is not completely per- 
manent most preparations are still perfect after at least thirty years and can at any 
time be remounted. Since the cuticular preparations are acellular sheets rather 
than pieces of tissue the number of stages from water to alcohol, alcohol to water, 
and water to glycerol can be reduced to a minimum, and each stage to a few seconds. 
The procedure adopted for preparing permanent mounts may be summarized : 
leaf fragments macerated in suitably diluted Jeffrey's Solution in an open watch- 
glass ; cuticular membranes washed in water, re-macerated in undiluted solution 
and re-washed in water if some cellular tissue remains, and transferred to a slide 
in 70% alcohol ; 70% alcohol replaced by Sudan IV solution for fifteen to thirty 
minutes ; shde irrigated with water, with 50% glycerol, and then with 100% 
glycerol ; and cuticular membranes finally mounted in glycerin-jeUy. 



24 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

Cuticular preparations can be kept indefinitely, before staining and mounting, 
in 70% alcohol ; placed in water they become excellent cultures of bacteria, fungi 
and protozoa (which feed on the adhering polysaccharides, not on the cutin) in a 
day or two. 

§5. Survey of cuticular characters 

The present section is a catalogue of the main characters which are present in 
cuticular preparations, together with some idea of the extent of their variation from 
species to species. In general the ranges of variation discussed indicate those 
normally encountered without examining plants of an extreme facies or with a very 
specialized adaptation, so that no mention is made of unusual extremes of structure 
or of features of very restricted occurrence. Any combination and number of these 
characters may be of diagnostic value in a particular group of plants. The first 
task is thus to survey the features found on cuticular membranes, and then to 
investigate the degree of variation each exhibits. After analysing the causes of 
variation the systematic value can be ascertained. Most of the comments made 
have been drawn from my experience in investigations of the three families 
mentioned previously, and, whenever relevant, examples are usually given from 
these groups. 

Such a general and, it is hoped, purely objective survey of cuticular characters 
has not been previously compiled, and thus the different terms which have been 
introduced into the literature are quite uncorrelated. In many cases, for instance, 
several descriptive terms exist for a single situation (e.g. cuticular flanges, teeth, 
pegs or ribs). When an acceptable and non-ambiguous descriptive term is already 
available this has been adopted, but in some cases terms had to be coined, or existing 
unsuitable ones replaced. All accepted terms are printed in italics, and it is hoped 
that this will encourage nomenclatural stabihty. 

Of the characters observed during microscopic examination of cuticular prepara- 
tions some are quite easily observed on whole leaves with an ordinary hand-lens, or 
with the naked eye. In these cases microscopic examination usually gives a more 
quantitative or precise assessment of the character, although some points (such as 
the colour or " texture " of the indumentum, e.g. silky, bristly or velvety) may be 
lost with this medium. Other characters are wholly microscopic in nature, and it is 
convenient here to treat the two groups separately. 

A. Macroscopic characters. 

The four features discussed here are treated only from the microscopic viewpoint, 
although preliminary study of a leaf with a hand-lens or low-power binocular 
microscope should always be undertaken. 

(i) Leaf venation. 

The venation of a leaf, second to its general shape, is frequently the most obvious 
character, and descriptions of all new plants contain (or should contain) a note of its 
organization. The type of venation is often typical of broad taxonomic groups. In 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



25 



general, dicotyledons have leaves (or leaflets) with a single central main vein 
which is a continuation of the petiole, if this is present. This is known as a midrib, 
costa or main vein. In some leaves two or more main veins may be present, 
although that situation is more characteristic of monocotyledons. Branching from 
the midrib are a number of lesser veins or nerves of successively smaller order. 
Those branching directly from the midrib may be termed lateral veins, and they are 
usually very numerous. However, a limited number, from two or three to a score 
or more, are generally much more conspicuous than the rest and are here described 
as major lateral veins, as opposed to the less distinct minor lateral veins. In most 
descriptions " lateral veins " usually refers to the former only. The number of 
major lateral veins, their spacing, whether they arise in pairs or independently on 
either side of the midrib, the angle at which they arise, the curvature along their 
length, and their method and position of ending are all taxonomic features. 



midrib 



minor lateral vein 
major lateral vein 



veinlet termination 
venule 




secondary vein 



tertiary vein 



Fig. 3. Terminology of leaf venation. 



Such a venation system is pinnate, but palmate systems also exist. In the latter 
case there are a number of main veins all arising from a single point and diverging 
outwards, and each usually has an accessory venous system resembling that of a 
pinnately veined leaf. 

Generally, arising from the lateral veins is a series of secondary veins, from these 
tertiary veins, and so on. Unlike the branching of a tree, however, the smallest 
veins, the venules or veinlets, in most cases do not possess terminations, but usually 
join up to form an extensive venule reticulum. The number of orders of veins between 
the midrib and the venules is usually one or two only, but it may be more. The 
term venule is reserved for the smallest veins with no directional tendency. What- 



26 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

ever the number of orders of veins, however, there is in these types of leaves usually 
a system of venules which emanates directly from the midrib, and almost always 
one from all other orders of veins. The venation system (Fig. 3) may thus be 
represented : 

Secondary Tertiary 

Midrib ^Lateral veins ^ veins ^veins, etc. 

, ^ ^ 

Venule reticulum 

The above systems of venation, terminated by a venule reticulum, are termed 
closed, but open systems, where a reticulum is not present, are found in some 
dicotyledons. 

The spaces in the venule reticulum are known as areolae, veinlet-islands, vein- 
islets or inter-reticular spaces. Their shape and size vary considerably and, although 
difficult to measure, are often the cause of the different appearance of various leaves. 
Levin (1929) found that the average " vein islet area " is of considerable taxonomic 
importance. 

Although most, or all, venules join up to form the venule reticulum there are often 
a number which end blindly, projecting into areolae, these being termed veinlet 
terminations (venule terminations). Hall & Melville (1951) calculated the " veinlet 
termination number " (number of veinlet terminations per square millimetre of 
leaf surface) and found it of diagnostic importance. It was stated to be uncorrelated 
with the number of areolae per square millimetre, and less variable over different 
parts of the same leaf and hence of greater systematic value. Later work by the 
same authors (1954) confirmed that the veinlet termination number was more 
constant than the number of areolae, but found that the two showed a slight positive 
correlation, and both a negative correlation with leaf size. Gupta (1961) improved 
Hall & Melville's methods by devising the " absolute vein-islet number " and 
" absolute veinlet termination number ", the product of the respective numbers 
and the leaf area. This, of course, cancels the variation due to leaf size and, although 
more laborious to calculate, should always be used when whole leaves are available. 
In certain disciplines, such as pharmacognosy and paleobotany, however, whole 
leaves are often not available ; in these cases the non-absolute numbers should be 
used with the greatest caution. 

Leaf venation is treated from a different aspect below, when its position as a 
cuticular character will become clear. 

(2) Distribution of trichomes. 

The very different degrees of pubescence shown by plants have given rise to a great 
variety of descriptive terms (e.g. hirsute, pubescent, pilose, ciliate, puberulent, 
etc.) which unfortunately have no universally recognized meaning. Using cuticular 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 27 

preparations the precise distribution and frequency of hairs and other trichomes 
can be measured, i.e. the individual trichomes can be counted and their positions 
noted. Taking into account the enormous variation that is encountered, only- 
differences of a considerable degree are likely to be of taxonomic value, and the 
extremes of a range must be noted. 

Precise data may thus sometimes replace such statements as " almost glabrous " 
or " sparsely pubescent ", but in general their compilation will not be worth the 
labour involved. Besides variation apparently uncorrelated with other factors the 
degree of pubescence is dependent more than most characters upon the environ- 
mental conditions and the age of the leaf. 

Of much more value than simple trichome frequency is the relative frequency of 
trichomes on different parts of the same leaf. As is well known, hairs are usually 
more abundant on the midrib and veins of a leaf, and these regions are usually the 
last to become glabrous as the leaf ages. However, this is by no means always the 
situation. In the Comhretum psidioides Welw. aggregate, for example, there are 
three taxa (treated by Exell (1961) as subspecies of C. psidioides) which differ only 
in their distribution of indumentum on the lower leaf surface : one taxon has a 
densely pubescent venule reticulum and sparsely pubescent to glabrous areolae 
(subsp. psidioides) ; the second a glabrous venule reticulum and tomentose areolae 
(subsp. kwinkiti (De Wild.) Exell) ; and the third a tomentose venule reticulum 
and areolae (subsp. dinteri (Schinz) Exell). Moreover, a further species in this 
aggregate, C. grandifolium F. Hoffm., differs not only in its larger leaves but also 
in having glabrous areolae and a sparsely pubescent to glabrous venule reticulum. 
This taxon, demonstrating the fourth combination, appears to be a further sub- 
species of C. psidioides. Dr. Exell and I have found that all four taxa belonging to 
this group have apparently no other constant differences, either macroscopic or 
microscopic, than the distribution of hairs. In many plant groups the distribution 
of hairs on the leaf margin is frequently also a diagnostic character. 

The broad type of trichome, e.g. sessile glands, stalked glands, branched hairs, 
peltate hairs, long straight hairs or short curly hairs, etc., is also a character often 
used without resorting to the use of a microscope, although this is not to be encour- 
aged. Small differences, often vital, may be not apparent with a hand-lens or 
binocular microscope, and very distinct taxa can easily be classed together. The 
key to the genera of Sapotaceae by Hutchinson & Dalziel (193 1), for instance, is 
completely misleading due to a mis-statement on the type of hair present. I have 
found that the hairs described as " simple (not medifixed) " (as opposed to medifixed, 
stalked and thus " T-shaped " hairs) are in fact medifixed but sessile. Such errors 
would have been avoided if a microscope had been employed. There have been 
many erroneous statements concerning the trichomes of the Combretaceae. A 
hand-lens should only be relied upon when two or more very different types of 
trichomes are being searched for (e.g. in Epilobium), or special distinctive trichomes 
are sought (e.g. in Urtica). 

As previously mentioned, characters of the texture or overall appearance (includ- 
ing colour) of the indumentum are lost in cuticular preparations. 



28 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

(3) Cork-warts and similar structures. 

Very frequently there are to be found on the cuticular preparations of leaves 
structures which are usually referable to wounds of one sort or another, but which, 
in some cases, represent cork-warts or related structures. Cork-warts are not 
abundant in nature, and they are usually constant in occurrence, so they are frequently 
useful in identification. In the mangrove genera of the Rhizophoraceae, for instance, 
they are present in all species of Rhizophora but in none of the other three genera. 
In West Africa, moreover, they are larger in R. mangle L. than in the other two species 
of that genus found there, and are said to be diagnostic in the field (Keay, 1953). 
They appear as small convex warts of corky tissue up to about 2 mm. across. 

Microscopically the position of a cork-wart on the cuticular membrane is marked 
by a circular or less regular hole surrounded by modified epidermal cells (Plate i a) . 
The latter are usually, though not always, thicker-walled and much smaller than the 
normal epidermal cells, and they are arranged in strict rows radiating for some 
distance from the hole. It appears as if the radiating cells are meristematic, being 
produced to repair the hole . These areas are thus very characteristic and conspicuous . 

Unfortunately three types of wounds are extremely similar to the cork-warts. 
Purely mechanical accidents are marked by variously shaped, but usually long and 
scar-like, areas, the surrounding cells being identical to those of the cork-warts. 
In many cases, presumably in the case of more ancient wounds, or perhaps when 
only part of the outer wall of the epidermis has been removed .or damaged, a very 
thin cuticular membrane, usually scarcely stainable, has grown over the wound, 
this being either amorphous or with very thin cell walls. The cells on these areas, 
if present, are usually normal-sized, although small cells would have been expected 
if they were meristematic. Insect punctures are also frequently found on leaves, 
these more often resembhng cork-warts than accidental injuries, and probably not 
being distinguishable from the former. They are of the same order of size as a small 
cork-wart, and usually smoothly circular in outline. They were also reported by 
Orgell (1955), who observed an identical structure. Moreover he found that they 
frequently became occupied by corky or gummy plugs, showing a very close analogy 
to cork-warts. The third type of wound which I have observed is a fungal attack, 
found in the form of hemispherical shining excrescences on both leaf surfaces of one 
specimen of Buchenavia kleinii Exell. The hole left by these fruiting bodies after 
maceration is surrounded by the radiating rows of small cells typical of cork-warts 
(Plate I B). 

Although cork is more or less resistant to the macerating solutions used in the 
present study, owing to its chemical similarity to cutin, it is not preserved on the 
cuticular preparations as the wart or plug is not connected to the cuticular membrane, 
but falls away during maceration. Where a wound has healed over, and occasionally 
a cork-wart has dropped out, however, the tissues may become covered with a thin 
cuticular membrane, as observed above. 

Thus, although wounds have no intrinsic taxonomic value, they are important 
as they closely resemble cork-warts, and thus render the latter considerably less 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 29 

useful in systematics. This equally applies to macroscopic and microscopic studies. 

Other types of structures found on leaves, although often not macroscopic 
characters, are sufhciently similar to cork-warts to be mentioned here. In many 
cases they may be referred to excretory glands (e.g. salt glands, chalk glands, sessile 
hydathodes, etc.), which may be macroscopically very conspicuous. The actual 
gland is frequently only very thinly covered with a cuticular membrane, and the 
adjacent cells may be radially arranged or otherwise modified. Holes may also be 
found in the cuticular membrane due to the presence of modified epidermal cells 
(e.g. calcilied or silicified cells) which have no cuticular covering, but these are usually 
surrounded by normal epidermal cells, as are the terminations of long idioblasts. 
Cells containing crystals frequently break through to the surface of the leaf, and 
these are often surrounded by radially arranged cells, as in wounds. In this 
connexion the minute pimples and translucent spots common on some leaves and 
often used by taxonomists may be mentioned. They are not epidermal characters : 
the former are due to abnormally large epidermal or subepidermal cells, usually 
containing crystals ; the latter to similar cells which interrupt the mesophyll. 
In fact in most cases the translucent spots correspond with small pimples, although 
either one may be present in the absence of the other, or rarely both exist together 
on the same leaf but are unconnected. 

Wound-like structures are frequently present on cuticular preparations in such a 
regular manner that they are almost certainly not wounds, but no macroscopic 
equivalent is visible and no microscopic subepidermal structures seem to be 
connected with them. In Terminalia catappa L., for example, all specimens 
examined showed ragged holes along the leaf margin (Plate i c), and T. chebula 
Retz. possesses similar areas there and on the lower epidermis (but not the upper). 
In T. bursarina F. Muell. similar areas are present in the major lateral vein axils 
on the upper epidermis, corresponding to the presence of domatia on the lower 
epidermis. They are caused by the growing of the domatia (see below) right 
through the leaf to the upper surface. 

Besides hydathodes, other water-secreting organs are the water-stomata, which 
are usually distinctly different from the normal stomata in some way (e.g. by their 
smaller or larger size). They are, of course, not macroscopic structures, although 
they are in some families united into groups which appear as small surface spots 
(Solereder, 1908). In the two Combretaceous mangrove genera, Lumnitzera and 
Laguncularia, large stomata which may be water-stomata are found scattered over 
the leaf surface. Usually they are surrounded by enormous subsidiary cells, and 
around the latter the epidermal cells may be radially arranged (Plate id). In 
extreme cases the radially arranged cells form an extensive area resembhng those 
around a wound or cork-wart, and indeed in some instances the large stoma has 
disappeared, leaving a hole. Solereder (1908), in fact, states that stomata may 
become underlaid with a development of cork which ultimately erupts to give rise 
to a cork-wart. Cork- warts have not been recorded from the Comhretaceae, 
however, and any connexion between the observed structures and cork is conjectural. 



30 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

In addition Lumnitzera possesses wound-like areas along the leaf margin (cf. 
Terminalia catappa) which appear to be distinct from those developing around the 
large stomata, and which are also unconnected with the shallow marginal pits 
thought to be domatia. 

Finally, the presence in many instances of thin-walled areas, often only a few cells 
in size, on the leaf epidermis may be noted. In some cases the component cells are 
much more or much less undulate-walled than the rest of the epidermal cells. 
Their origin is unknown. 

It may thus be seen that the presence of " spots ", " pimples " or " warts " on 
leaves may be due to any one of a whole range of epidermal (or subepidermal) 
structures, which only a microscopic examination will determine. Their study is, 
however, hampered by the presence of various types of wounds and punctures of 
external cause which may extremely closely resemble the structural features and 
thus limit the determinative value of the latter. 



(4) Domatia. 

The term domaUum is used to denote a small cavity of very varying form, found on 
almost any vegetative part of the plant, which is inhabited, or thought to be 
inhabited, by some type of animal. Domatia have been almost completely neglected 
from a taxonomic viewpoint in the past, and so limited is our knowledge of them 
that they are treated rather fully here. It will be seen that on occasions they may 
be of great diagnostic significance. 

Domatia vary from cavernous passages in rhizomes, stems or petioles to minute 
pits or hair-tufts on the under surfaces of leaves. The former are usually occupied 
by ants, and the literature concerning these so-called myrmecodomatia is fairly 
extensive (see Uphof, 1942). The smaller domatia, which may be up to about 
4 mm. in diameter, are often said to be occupied by mites or similar arthropods, 
although good evidence for this is usually lacking. They are, however, called 
acarodomatia, a term suggested by Lundstrom (1887), and the plants on which 
they occur acarophytes or acarophilous plants. Elliot (191 1) stated that the mites 
could be seen scurrying away from the appropriate regions of the leaves of many 
common English trees when the former were disturbed with a needle, and Mani (1964) 
reported that acarodomatia may be inhabited by mites such as species of Tarsonemus. 

The literature relevant to acarodomatia is remarkably scanty, the only important 
papers being a biological study by Lundstrom (1887) and a review by Penzig & 
Chiabrera (1903). The latter account is excellent from many aspects, and thus any 
sort of systematic review is unnecessary here. The authors reported acarodomatia 
from about 425 species and 175 genera in 44 families, all woody dicotyledons. They 
are most abundant in the Ruhiaceae, in which Penzig & Chiabrera recorded 54 
acarophilous genera. As these authors predicted, since 1903 many additional 
genera have been shown to possess acarodomatia, but few are of very special note. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 31 

An exception is Dioscorea, which is the only monocotyledonous genus and also the 
only herbaceous one known to possess acarodomatia. 

Apart from the fact that they are more or less confined to woody dicotyledons 
there is no connexion whatsoever between the type of leaf and the presence or type 
of domatia. They do perhaps tend to be rarer on leaves with inconspicuous 
venation, but they are not absent from these and any generalization would have a 
number of exceptions. 

The commonest types of acarodomatia are small pockets or pits in the axils of 
the main veins with the midrib on the lower leaf surface. Since the whole domatium, 
whatever its shape, is lined with epidermis throughout, cuticular preparations 
provide excellent transparent models of the domatia and afford the best method of 
studying their anatomy. The preparations can be studied both with a high-power 
and a binocular microscope. The domatia are very conspicuous on the cuticular 
preparations since, where there is an extended piece of leaf tissue overarching the 
main bulk, the cuticular membrane overlaps and so is three layers in thickness. 

There is in all probability no system of classification that will accommodate all 
types and grades of domatium. The usefulness of some sort of classification is, 
however, obvious, and the one which I use to cover leaf domatia has proved 
relatively satisfactory. Two main types of domatium are recognized (Fig. 4) : 

Marsupiform [pocket-shaped) domatium [domatium marsupiforme) (Plate 2 a, b). 

All grades, from an extremely highly developed " pocket " to a situation which is 
scarcely able to be called a domatium exist. They are found in the axils of the 
major lateral veins with the midrib, or more rarely of secondary veins with major 
lateral veins. In a leaf with very conspicuously raised veins a cuticular preparation 
will always show folding of the epidermis of the veins over that of the areolae, and 
this effect is of course doubled at a vein branch, especially the branchings from the 
midrib. Leaves of Buchenavia reticulata Eichl., for instance, show this to a marked 
degree. In some leaves this folding is only evident in the major lateral vein axils, 
and here a V-shaped suggestion of a domatium is to be seen, as, for example, in 
Terminalia mollis Laws. Penzig & Chiabrera (1903) classed this species as 
domatium-bearing, but the case is debatable. In the genus Terminalia, as probably 
in many others, a whole series of types from that of T. mollis to the most complex 
pocket-shaped domatium is to be seen. This gradation can be visualized as the 
gradual filling-in of the notch of the V to form a V-shaped triangular pocket where 
all traces of the " arms " of the V have disappeared. These latter types bear no 
relation at all to the conspicuousness of the veins, often being found on leaves with 
an almost planar venation. The basal point of the triangle points towards the vein 
axil, and the mouth of the domatium, forming one side of the triangle, away from it. 
Thus the depth of the domatial cavity, which lies between the leaf surface and the 
triangular area of tissue, is parallel to the leaf surface. The mouth of the pocket- 
shaped domatium is therefore at its broadest and most distal point. 



32 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



ll'll'if 'l' ^ ./ 



' 5 




Fig. 4. Types of domatia : i, primary-axillary lebetiform. 2, primary-axillary 
marsupiform. 3, 4, vertical sections of i and 2 respectively, at angle bisecting the 
vein axil. 5, primary-axillary hair-tuft. 6, primary-axillary pubescent-marsupiform. 
7, marginal lebetiform, in section {Lumnitzera) . 8, baso-laminar revolute (Strephonema). 
9, baso-laminar revolute (Dioscorea) . 10, twinned supernumerary lebetiform [Terminalia 
catappa) . 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 33 

Lehetiform [howl-shaped) domatium [domatium lehetiforme) (Plate 2 c, d). 

As with the preceding type all grades of bowl-shaped domatia are to be found, the 
simplest of which might be visualized as a shallow depression in the leaf surface. 
Such shallow depressions are found in Lumnitzera, on the leaf margin, and have been 
referred to as domatia by Backer {fide van Steenis, in litt.). In most cases, however, 
bowl-shaped domatia are found in the axils of veins. In Terminalia plagata Merr. 
well-developed domatia are present in the major lateral vein axils, but in the 
secondary vein axils reduced forms, the shallow depressions visualized above, are 
to be found. A series may again be traced where the opening of the hollow becomes 
smaller, and/or the hollow itself becomes deeper and wider, the genus Terminalia 
showing all types of intermediates. As noted by Penzig & Chiabrera (1903), in 
some cases (e.g. T. microcarpa Decne.) the pore of the domatium may be at the level 
of the rest of the leaf surface, the pit being entirely internal, whereas in other cases 
(e.g. T. catappa) the pit is only partially sunken into the leaf tissue, the pore being at 
the apex of a small dome. In the best-developed bowl-shaped domatia the pore is 
extremely small in proportion to the large pit beneath (e.g. T. microcarpa). In 
the present type the depth of the domatium is thus at right-angles to the leaf surface. 
The pore is usually in the centre of the domatium as viewed from above, but it may 
be placed elsewhere. In some abnormal forms of domatium in T. catappa there are 
two pores. The pore may be circular, elliptic or slit-like : in the last case it is most 
frequently found on elongated domatia, the elongation of both domatium and pore 
being parallel to the lateral vein. 

The above two types of domatium are nearly always distinct, and it is suggested 
that they arose independently, from the overarching veins and the shallow hollow 
respectively. Intermediates, although rare, are apparently present, but they are 
thought more likely to represent the points of convergence than of divergence. As 
shown above, increasing stages in complexity of both types can be traced which are 
difficult to interrelate. In apparently intermediate types (e.g. Terminalia moluccana 
Lam.) the bowl-shaped domatium has a large pore which is distal to the vein axil, and 
the cuticular membrane is often not trebled at its distal edge. The trebled membrane 
thus forms a crescent rather than a circle, and moreover the depth of the domatium 
appears to be parallel to the leaf surface, although it may not be so. This type 
obviously closely resembles the pocket-shaped domatium. The latter may also 
approach the intermediate situation by the narrowing of the opening. The 
domatium of Buchenavia fanshawei Exell & Maguire, which is classed as the bowl- 
shaped type, may possibly be a modified pocket-shaped type as all other species of 
Buchenavia which possess domatia have the latter type. 

Two other types of domatium may be recognized : the first often very well 
defined ; the second of a highly dubious nature. 

Revolute {rolled-margin) domatium {domatium revolutum). 

In this type the leaf margin is inroUed, or folded under like the flap of an envelope. 
Inrolled domatia are found in such plants as Hevea brasiliensis (Kunth) Muell. Arg. 



34 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

(Penzig & Chiabrera, 1903) and Strephonema spp. (De Wildeman, 1923 ; confirmed 
by my own examination). The inrolling is present at the base of the lamina, just 
above the petiole on either side. The degree of inrolling varies considerably, and 
the status of the domatium is thus dubious. In the flap-like domatia a specialized 
portion of the lamina is folded under ; if this is folded back it forms a regularly 
shaped projection from the leaf margin. It was found in dicotyledons (e.g. Ilex sp.) 
by Lundstrom (1887) and later in Dioscorea by De Wildeman (1904). The latter 
author described the structure in D. smilacifoUa De Wild., in a new species 
{D. acarophyta De Wild.), and subsequently, from 1912 to 1914, in ten other new 
taxa. Burkill (1939) showed that these twelve represented three species only : 
D. smilacifoUa, D. minutiflora Engler and D. pynaertii De Wild., but he discovered 
domatia in a fourth closely related species, D. praehensilis Benth. 

Hair-tuft domatium [domatium fasciculatum) . 

The vein axils of the otherwise sparsely pubescent or glabrous leaves of a large 
number of species of trees may be densely pubescent or pilose, and these tufts of 
hairs have been considered to be domatia by Lundstrom (1887) and others. If 
mites do take refuge in them this may be accepted. Amongst British trees Tilia, 
Alnus and Corylus are examples. In this type there is no modification of leaf tissue 
other than the development of the hairs. Although many trees have leaves which 
are at first densely pubescent and later scarcely so, and the last areas to lose the 
indumentum are the veins of the lower leaf surface, the hair-tuft domatia are often 
well defined, even if their association is dubious. They are frequently found in 
connexion with the slightest development of a pocket-shaped domatium, 

Lundstrom (1887) recognized more or less the same four types of domatia as I do, 
although a much smaller range of forms was considered. Penzig & Chiabrera 
(1903) closely followed Lundstrom, but described six types. The " hollow "-shaped 
{fossette) and " pocket "-shaped [tasche or borsette) domatia were each divided into 
two categories according to the presence or absence of hairs. It is not, however, 
advisable to distinguish between glabrous and pubescent types because the difference 
is apparently of no consequence and every type of intermediate is to be found. 
Penzig & Chiabrera appear to separate the " hollows " from the " pockets " in that 
the former are sunken into the leaf tissue with their openings on the level of the leaf 
surface, whilst the latter (which may be either bowl-shaped or pocket-shaped in 
my system) are raised above the leaf surface. Intermediates are, however, as 
common as either extreme, and the classification is considered inferior to that used 
in the present study. It so happens that Lundstrom's scheme does not commit 
itself to agreement with either Penzig & Chiabrera's system or mine, since all his 
" hollows " were in fact wholly sunken into the leaf, no raised hollows (as in 
Terminalia catappa) being noted. 

The latinized terms have been given for the convenience of concise description. 
The position of the domatia can also be described by the following terms : marginal, 
for those at the leaf margin (e.g. Lumnitzera) ; primary-axillary, for those in the 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 35 

major lateral vein axils (the most common type) ; secondary-axillary , for those in 
the secondary or lesser vein axils (e.g. Tilia spp., Terminalia oreadum Diels and 
T. plagata) ; haso-laminar , for those on either side of the lamina base (e.g. Quercus 
rohur L., Strephonema spp. and Dioscorea spp.) ; extra-axillary, for those just below- 
each major lateral vein axil (e.g. Rudgea lanceolata Benth.) ; and supernumerary, 
for those not associated with any particular part of the lamina (e.g. Terminalia 
catappa, where they are found scattered up the side of the midrib apart from the 
vein axils ; and T. grandiflora Benth. and Conocarpus spp., where they are found 
very sparsely scattered over the lower leaf surface). They are occasionally also 
found elsewhere : for example in Schinus terehinthifolia Raddi domatia revoluta are 
found on leafy wings to the rhachis of the pinnate leaves ; and in one or two genera 
they have been reported from young twigs and petioles. 

In most cases the epidermis lining and around the domatia is little or not modified. 
In the axils of major lateral veins there is usually a patch of small straight-walled 
isodiametric cells, and when a primary-axillary domatium is present this area is 
merely enlarged, and this type of epidermis lines the domatium. In some cases, 
however, the cells do differ in some way from those surrounding the domatium, 
usually being smaller. In Terminalia catappa the adjacent cells are often arranged 
in radial files for a short distance. In Conocarpus erectus L. var. erectus the epidermis 
is absolutely identical in, near and away from the domatia, even to the extent of 
the possession of stomata. In C. erectus var. sericeus Forsstr. ex DC. the epidermis 
is very densely pubescent, but that lining the domatia is more or less glabrous, 
as in var. erectus. 

In many cases the domatia are associated with an abnormal amount of hair 
development in, on or around the actual structure. In some species, e.g. Thiloa 
gracilis (Schott) Eichl., the hairs are of a peculiar type not found elsewhere on the 
leaf. In a few species gland-like structures are to be found in the domatium. These 
were noted in Anacardium by Lundstrom (1887), and in the Comhretaceae they have 
been found in Buchenavia parvifolia Ducke and Terminalia archboldiana Exell, 
though not very frequently. The nature and function of their secretion, if any, 
are unknown. It is just possible that the deeply sunken glands found on the leaves of 
Laguncularia are homologous with domatia, but this is not likely as they are found 
on both epidermides and have an extremely narrow slit-like opening at the surface. 

The examples cited above clearly show that domatia may have a considerable 
value in taxonomy, although they have rarely been used. An extreme case is per- 
haps that of Buchenavia fanshawei which has well-developed domatia lebetiformia. 
Of the other 22 species examined in this genus 13 have domatia marsupiformia and 
nine lack domatia altogether. Thus much use of this fact was made in constructing 
a key to the genus (Exell & Stace, 1963). In the Comhretaceae domatia revoluta 
are found in aU six species of Strephonema, which is usually treated as a separate 
subfamily from the other genera, all of which lack these domatia. The remaining 
genera belong to three groups, of which two contain eight genera each. In one of 
these groups (Terminalieae) four genera possess domatia (three lebetiform and three 



36 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

marsupiform) ; whilst in the other [Combreteae) all eight genera possess marsupi- 
form domatia only. The third group {Laguncularieae) consists of three genera, 
of which one possesses peculiar marginal lebetiform domatia. Thus in the 
Combretaceae the occurrence of the various types of domatia to some extent agrees 
with the generally accepted subdivision of the family. This is not the usual situa- 
tion, however. Penzig & Chiabrera (1903) recorded all six of their types from the 
Rubiaceae, and four from several genera, and in general domatia are useful in 
identification at the specific level only. 

There is no reason not to believe that all specimens of a domatium-bearing species 
do possess domatia, although these may be sparser in some cases and may be lacking 
altogether on some parts of the plant such as the sucker shoots. Bloembergen 
{fide van Steenis, in litt.) found that Alangium chinense (Lour.) Harms and A. kurzii 
Craib possess domatia in most of Asia, but not in Malaysia, but, as this situation 
has never been reported elsewhere, the absence of domatia in Malaysia or the identity 
of the species in the two areas needs careful checking. In the genus Terminalia, 
of which I investigated 160 species, no such geographical variation was noted. For 
example T. catappa, a native of Asia and Australasia, possesses very typical domatia 
whether in its native countries or planted in Africa or America ; and T. arjuna 
(Roxb.) Wight & Arn., a native of India, lacks them there or when planted elsewhere 
in Asia, Africa or America. A different type of geographical variation, however, 
was discovered in Terminalia. Of the 160 species studied 50 (31%) possess domatia, 
of which 16 (32%) have domatia marsupiformia and 34 (68%) domatia lebetiformia. 
However, in Australasia 21 (62%) of the 34 examined species possess domatia, and 
the figures for other continents are : Malaysia 19 (58%) out of 33 ; Asia 4 (13%) 
out of 34 ; Africa i (3%) out of 39 ; and America 7 (23%) out of 30. Thus as one 
travels east or west from the islands between Australia and Malaysia the percentage 
of domatium-bearing species drops sharply. The proportion of the two types of 
domatia does not significantly differ from continent to continent. 

Extremely little is known about the formation or function of acarodomatia, but 
despite suggestions to the contrary it is obvious that mites play no part in their 
formation. It is more likely that the mites inhabiting the domatia (if this is a regular 
occurrence) are in a purely chance association. It is possible that they may enlarge 
the domatia somewhat once they have taken possession of them, and this may be 
the cause of the deep domatia in Terminalia bursarina which sometimes reach the 
upper epidermis. There is no necessity, however, to consider that this is not a natural 
plant process. Lundstrom (1887) claimed that the domatia of Psychotria, Coprosma 
and Rhamnus developed in the absence of mites, but, although he was doubtless 
correct, his experiments have never been repeated under the exacting conditions 
necessary to exclude all mites. Mani (1964) also stated that the mites did not 
participate in domatium formation. Bailey (1924) showed that American myrmeco- 
domatia developed in the absence of ants. Whilst it is true that seedlings (and often 
suckers) lack domatia, these developing only at a later stage, seedling and sucker 
leaves often differ in many ways from the other leaves, and van Steenis (1953) states 
that in Nothofagus domatia develop by the young sapling stage. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 37 

Good evidence for the non-participation of mites in domatium formation is that 
domatia are definitely formed in predisposed places : e.g. at the bases of the laminae 
in Dioscorea, or just below each major lateral vein axil in Rudgea. If the domatia 
were formed by mites they would not be so distributed, and neither would they 
be so constant in their presence (on all individuals in all geographical areas) and 
structure. In the case of primary-axillary domatia these are not always found in 
all major lateral vein axils. They are always better-developed towards the lamina 
base, but in some species, e.g. Terminalia bursarina, T. grandiflora and Nothofagus 
menziesii Oerst., they are found only in the lowest axil on each side of the midrib, 
or in the latter two sometimes also in the next axil. Furthermore, only certain 
species in any one genus may possess domatia, e.g. only two species out of about 
40 in Nothofagus (van Steenis, 1953) and four out of several hundred in Dioscorea 
(Burkill, 1939). The domatia are often very hairy, and since many mite-produced 
galls also possess this character, definitely as a direct result of the animals' activities, 
it has been suggested that pubescent domatia are also formed in this way. There 
is no reason to believe this, however, and in most Comhretaceae, at least, the pubescent 
domatia (often the only pubescent parts of the leaf) are relics of a wholly pubescent 
juvenile state, and the hairs may wear off even here at senescence. 

The very young leaf develops domatia marsupiformia by an extension of the 
pocket tissue over the vein axil, but, if domatia lebetiformia evolved in the manner 
previously suggested, ontogeny does not repeat phylogeny as the pits do not appear 
to develop by an extension of tissue encroaching over a hollow. In the cases 
examined (especially Conocarpus erectus) the pit forms internally, and at first is 
only marked by a primary axillary pimple. Later a pore develops in the top of 
the pimple, and this enlarges to a varying degree. The hollow appears to develop 
internally. Different stages may be seen on one leaf, development being acropetal. 
This is an important argument against any participation by mites in domatium 
formation. Holtermann (1893) , however, seems to suggest that the pits of Conocarpus 
do develop by means of an epidermal intucking, and not as I have observed. 

The situation regarding domatia has been discussed at some length, although 
numerous other points could have been included, since little is known of them in 
general, and they have been largely neglected by taxonomists. With regard to 
their position and structure they may be very useful in identification, but their 
biological significance is highly obscure. They are doubtless often used as a shelter 
by mites and other similar minute arthropods, but in the great majority of cases 
no animals have been connected with them. This is in marked contrast to the case 
regarding myrmecodomatia. The prefix " acaro- " is thus best rejected and replaced 
by one suggestive of the organ on which the domatia are found, e.g. phyllodomatia. 

B. Microscopic characters. 
(i) Epidermal cells. 

There are a large number of directions of variation to be found with respect to the 
epidermal cells, even omitting the several characters which are not apparent from 
a purely surface examination, but they have rarely been used by taxonomists. 



38 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

In certain regions the epidermal cells are modified, notably over the veins and leaf 
margin and around the stomata and trichome-bases. This section deals only with 
the areas not so modified. 

Although the shape, size and other characters of the epidermal cells exhibit a wide 
range of genotypic variation, which in many cases has definite taxonomic applica- 
tion, the cells certainly show an extreme degree of phenotypic variation which is 
elsewhere only paralleled by characters such as pubescence and petal colour, etc. 

One of the most obvious features of the epidermal cells is their shape ; this is 
often quite different on the two epidermides. This fact is often of systematic 
importance, although Odell (1932) decided that it indicated that cell shape was 
" of no diagnostic value ". The cells are usually isodiametric, i.e. more or less the 
same in length and width, with any scattered elongated cells randomly orientated. 
In some dicotyledon leaves, however, as well as in many groups of monocotyledons, 
the epidermal cells are predominantly elongated. This feature usually occurs in 
narrow leaves, the cells lying parallel to, or less often at right-angles to, the leaf axis. 
In cases of isodiametric-celled epidermides some species have a characteristically 
low (e.g. Guiera senegalensis J. F. Gmel., Plate 4 d) or high number of walls to each 
cell, this factor markedly altering the appearance of the epidermis. 

In some species the epidermal cells are arranged in small groups separated by 
thicker than normal cell walls, where it appears that an originally single cell has 
retained its identity after many subsequent divisions. Species of Avicennia show 
this feature well. A related phenomenon is the anticlinal division of the epidermal 
cells, where cells are divided by one or more thin straight walls which cut a normal- 
sized cell into two or more parts. Such secondary division (i.e. occurring after 
normal epidermal cell division has ceased, and not being accompanied by an increase 
in size) is often taxonomically important. In the genus Laguncularia, for example, 
it is absolutely constant, and it is not well developed elsewhere in the family 
Comhretaceae. 

The term " wall " is here used as an abbreviation for the apparent cell wall 
produced by the cuticular flanges, and it should always be borne in mind that the 
actual cell wall was present for some distance on either side of these. For this 
reason some workers prefer the term " cell outline ". 

The anticlinal walls of the epidermal cells are either straight, curved or variously 
undulate or sinuate (cf . Plate 3 a, b) . This is undoubtedly of taxonomic use in some 
instances : it has been used as a major character in Lemna and is useful in many 
groups of Combretum. For some unknown reason straight-walled epidermal cells 
are commoner in xeromorphic plants than in mesomorphic ones, which typically 
have undulate cell walls. The relative advantages and disadvantages of undulate 
walls are unknown, but Wylie (1943), who attributes a conductive function to the 
epidermis, suggests that their advantage is to increase the area of contact between 
adjacent cells. If this is so it is surprising that undulate walls are not commoner 
in xerophytes. The type of undulation is very various ; the " waves " may be 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



39 



most easily measured in exactly the same terms as light waves, i.e. in the three 
variables frequency, amplitude and wave-length. In the present study the frequency 
was usually expressed in terms of wave-lengths per wall, but in some cases it seems 
that the number of peaks (i.e. half wave-lengths) per cell may be a less variable 
value. There may be considerable or very little variation in the degree of undulation 
on one leaf, but the undulation often markedly decreases towards the leaf margin 
and veins. It will be seen from Fig. 5 that the three variables mentioned above 
aptly express all the characters of the undulations but one, i.e. the shape of the waves. 
These may usually be described as U-shaped, V-shaped or Q-shaped. In some 
species peculiar shapes of undulation are found with ornamentations, in the form of 
knobs, ridges or T-shaped thickenings, on the outer side of the crest of each wave. 
Curved epidermal cell walls may be described as having half a wave-length. 




v/V^ 




Frequency 
10 

10 
20 
20 



Wave-length 



Amplitude 

2y 



2-5 



2-5 



2x/ 



2y 



2y 



2y 



0-5 2x y 

Fig. 5. Types of cell-wall undulation. 

The exact causes of epidermal cell-wall undulation are unknown, the two main 
theories involving the suggestions that it is due to the tensions set up between the 
mesophyll and the epidermis, and that it is caused by the method of hardening 
of the differentiating cuticular membrane (Watson, 1942). It has been found 
that whereas many leaves have the epidermal cells undulate-walled throughout 



40 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

the height of the cells, some have only the outer part of the wall undulate, and thus 
the second hypothesis is by far the more likely. Sifton (1963) has provided further 
evidence supporting Watson's theory. Amongst species examined in the present 
study the situation where only the outer part of the wall is undulate is very con- 
spicuous in Ramatuella. Watson, however, was able to produce either condition in 
mature ivy leaves by varying the illumination during development, so that the 
character is probably rarely of taxonomic use. In some species with straight walls 
the isolated cuticular membranes show the tips of the cuticular flanges to be irregu- 
larly undulate, but this is probably due to the effect of removing the epidermal 
cells. It is especially noticeable in Conocarpus eredus. 

As mentioned above, the cell shape is often different on the two epidermides, the 
chief varying factor being the degree of undulation. In almost all undulate-walled 
leaves the amount of undulation is greater on the lower than on the upper epidermis. 
In many cases the former is markedly undulate-walled and the latter straight-walled. 
Odell's (1932) remark that a few leaves of Acer pseudoplatanus L. may show the 
reverse of this situation, which the bulk of the leaves exhibit, needs to be checked. 
vSome species, however, do possess a markedly more undulate-walled upper epidermis 
than lower, e.g. Terminalia microcarpa, T. plagata and others. 

Epidermal cells vary tremendously in size, average widths being about 20-30 |i. 
Even on typically large-celled epidermides, however, small cells are usually to be 
found scattered. Some epidermides, in fact, are characterized by their extremely 
variable-sized cells. That cell size might be of taxonomic value is indicated by the 
knowledge that polyploids usually have larger ceUs than diploids, although this is 
better shown by the stomata than other epidermal cells. Solereder (1908) mentions 
the fact that certain species are characterized by a particularly small- or large-celled 
epidermis, and this has been noted in the present study. Avicennia marina var. 
rumphiana (Hallier f.) Bakh., for example, has a notably smaller-celled epidermis 
than other varieties of this species. Lumnitzera littorea (Jack) J. O. Voigt has a 
smaller average cell size than has L. racemosa Willd., although some overlap occurs 
and this character, like all others, fails to separate the two species on surface cuticular 
characters in all cases. The cell size is frequently different on the two epidermides, as 
is cell shape. Solereder (1908) states that in general the upper epidermis is composed 
of larger cells than the lower, but the opposite was found in the present study. 
Characters such as the comparative sizes of parts of the same leaf are extremely 
useful as the different parts are all equally affected by external conditions, e.g. 
the presence of upper epidermal cells twice as large as the lower epidermal cells is a 
better diagnostic character than the presence of upper epidermal cells of an average 
length of 40 [X. The cell size usually decreases towards the leaf margin and apex. 

Although the boundary between the cuticular membrane and the cellulose cell 
wall is usually a smooth line, interrupted only by the cuticular flanges, in some species 
regular cellulose pegs protrude outwards into the cuticular membrane, which becomes 
very thin in these regions. This may or may not be accompanied by an extension 
outwards of the cell lumen in the corresponding position. In either case small 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 41 

elliptic or circular areas are seen on the cuticular membranes where the thin 
membrane is less deeply stained. Such areas are very well developed in Combretum 
obanense (Baker f.) Hutch. & Dalziel and certain of its relatives, but in all other 
species of this genus they are unknown. The thin areas are frequently present in 
the bulge of each undulation, and in C. obanense they are so conspicuous that the 
normal differentiation between cuticular flanges and non-flange areas is entirely 
masked, except on the veins. The epidermis in this case appears to be composed 
of minute cells about 5-10 [i across, with extremely thick walls. Other situations 
have been described where the outer epidermal cells contain large pores, the cell 
contents having been lost. The origin of the pores in the cuticular preparations of 
Cinnamomum obtained by Marlier-Spirlet (1945) is obscure. 

The surface of each epidermal cell may be flat or convex, or bear a small conical 
process. The latter is known as a papilla, and the epidermides or cells which bear 
papillae are said to be papillate, but unless the papillae exceed a certain length they 
are not seen on surface preparations of cuticular membranes. They may, however, 
be viewed on folded-over pieces of membrane. All gradations from simple conical 
papillae to long structures better described as hairs exist, and if they are long enough 
to become folded over on cuticular preparations they are visible in surface views. 
They are often considered to be taxonomically important. The papillae are usually 
centrally placed on the epidermal cell, but they may be found on one corner of 
each cell, as in Combretum zenkeri Engler & Diels. Elsewhere in this large genus 
papillae have been recorded in less than half a dozen species, e.g. C. lanceolatum 
Pohl ex Eichl., and in all cases are of the central type. 

(2) Leaf venation. 

The epidermal cells above and below the veins are usually variously modified, so 
that leaf venation is a cuticular character. Characters of the venation which are 
microscopic alone, as opposed to the type of venation (which is a macroscopic 
character), are the extent and nature of the modification of the epidermal cells. 

The midrib and the lateral, secondary, etc. veins are usually enclosed or capped by 
sclerenchyma, which may or may not reach either epidermis. In any case the 
epidermis above and below these veins is usually markedly modified, in the majority 
of cases the cells being elongated parallel to the veins. These elongated cells are 
usually rectangular, and, even if the epidermal cells of the areolae have undulate 
walls, straight- walled. In general the more xeromorphic a leaf the less conspicuous 
are the veins on the epidermis. Usually the degree of modification of the epidermis 
is proportional to the prominence of the veins, hence there tends to be a much 
greater modification on the lower than on the upper epidermis, although this is not 
always so. In most cases a more modified epidermis has more narrowed and 
elongated cells over the veins. The ends of the cells may be truncate, oblique or 
almost pointed and interlocking (prosenchymatous) ; and the cells arranged end to 
end in numerous parallel rows, or less regularly placed. In many cases, especially 
when the midrib is well developed on the epidermis but the cells are not very narrow, 
the cells are secondarily subdivided at right-angles to the midrib, and in these cases 



42 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

the cells may be squarish or even broader than long. Some leaves, most frequently 
rather xeromorphic leaves with a prominent midrib but rather inconspicuous veins 
and venules, have the epidermis modified over the former into strict rows of broader- 
than-long cells, the primary and secondary cell divisions being indistinguishable 
because all the cuticular flanges are equally thick. Examples found in the Combre- 
taceae are Strephonema spp., Combretum demeusei De Wild., Terminalia laxiflora 
Engler and many others. This effect is almost always more greatly developed on 
the upper epidermis, and furthermore the cells of the veins of lesser prominence 
show it to a progressively lesser degree. 

The lesser veins and venules showing no directional tendencies have a different 
anatomical structure. Above and below, or below only, many of them are bounded 
by a dorsiventral extension, the vein-extension or vein-rib, which consists of living 
cells elongated parallel to the venules and abutting on to the epidermis below and 
often also above. Wylie (1943) considers that these, as well as the epidermis, 
constitute a conductive system supplementary to the venous system. When a 
venule is connected to the epidermis in this way the latter is modified. Thus it 
may be seen that, as with the major veins and the midrib, the upper epidermis is 
modified to an equal or lesser extent than the lower epidermis over the lesser veins 
and venules. The modification of the cells over the venules (Plate 5 b) is similar 
to but much less developed than that over the midrib, etc. The cells are generally 
elongated, to varying extents, although secondary division of them is rare or absent. 
When the epidermis is mostly composed of undulate-walled cells the venule cells 
are usually straight-walled, but sometimes they are slightly, or even equally, 
undulate- walled. In the latter case only their elongation and, on the lower epider- 
mis, the lack of stomata make them distinguishable. 

Thus every intermediate between the very long and narrow midrib cells to the 
scarcely modified venule cells exists. The cells in the centre of each vein are mostly 
greatly modified, the edges of the veins gradually or abruptly merging into the 
normal epidermal cells. The most modified midribs are several hundreds of cells 
wide, the smallest venules only one or two. In an average mesomorphic leaf such as 
those of many species of Combretaceae the following degree of modification is seen : 
midrib composed of about 50 parallel rows of narrow elongated straight-walled cells ; 
major lateral veins about half as wide, the cells shorter and wider ; secondary veins 
about 10-15 cells wide, the cells only about twice as long as broad ; lesser veins 
about 5-10 cells wide ; venules forming venule reticulum about 2-3 cells wide, 
the cells little longer than broad, many with slightly undulate walls. Variation 
in one extreme is shown by strongly xeromorphic leaves, e.g. of Laguncularia, 
where only the midrib is modified on the epidermis, and that not conspicuously ; 
and in the other extreme by species with strongly reticulate leaves, e.g. Buchenavia 
reticulata, where the venules forming the reticulum are 5 or more cells wide, the cells 
being straight-walled and somewhat narrow and elongated. In all cases the lower 
epidermis is more greatly modified than the upper, so that the venation of the upper 
epidermis of a leaf would resemble that of the lower epidermis of a more xeromorphic 
leaf, and so on. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 43 

From the preceding paragraphs it can be seen that the three variables mentioned 
(i.e. the venation pattern ; the degree to which the venation system is recognizable 
on the epidermis ; and the type of modification of the cells) are completely indepen- 
dent of each other. Because of this, caution must be exercised in interpreting the 
extent of the venation system seen on the epidermis as that to be found in the leaf : 
it is frequently not so. The smallest veins represented on an epidermis, whether 
they are in fact the smallest venules of the leaf or of a much larger order, tend to be 
of the 2-3-wide slightly elongated-cell type. Thus a venule reticulum with areolae 
3000 [X across seen on an epidermis might represent this structure in the leaf, or 
smaller venules may be present but not differentiated on the epidermis. According 
to the excellent figures of Wylie (1943), and to my own observation, the ultimate 
veinlet terminations rarely have vein-extensions, and are thus rarely visible on an 
epidermis. Apparent veinlet terminations seen on the epidermis are due to venules 
which are not represented in full. 

Although the vein-like structures seen on the epidermis are of course only modified 
epidermal cells, they are for convenience referred to as midribs, veins and venules, 
etc. Thus the " upper epidermal midrib " or " lower epidermal venules " may be 
referred to in descriptions. The unmodified epidermal islands within the venule 
reticulum may likewise be referred to as the areolae. In cases where the venules 
are not well represented on the epidermis, and the venation system is thus apparently 
open rather than closed, the unmodified epidermal cells may be referred to as the 
non-venous areas. In general the differentiation of the veins on the epidermis is 
subject to rather more phenotypic variation than is the organization of the actual 
veins themselves, calling for a greater measure of caution in their taxonomic use. 

(3) Leaf margin. 

The epidermis is continuous over the leaf margin, connecting the upper and lower 
epidermides, but its cells are modified at these positions, providing useful taxonomic 
characters. Towards the margin the cells of both epidermides are frequently smaller 
and thicker-walled, and if the normal cells are undulate-walled the undulations 
become less pronounced. Thus the marginal cells either abruptly or gradually 
merge into the normal epidermal cells. 

The width of the margin, of course, is partly dependent upon the thickness of 
the leaf, and may be of diagnostic value. In most mesomorphic leaves the margin 
is not very wide, often not exceeding 5 or 6 cells across. The cells are in most cases 
rectangular, straight-walled and variously elongated (Plate 3 c) . Secondary division 
of these cells may be apparent, giving rise to squarish or broader-than-long cells. 
The margin is thus frequently very similar to a venule, and may be indistinguishable 
but for its continuous straight appearance and the absence of other venules branching 
from it. 

In certain cases, especially in xeromorphic leaves, other types of leaf margin are 
to be found. One of the most common of these is composed of small isodiametric 
cells, frequently with very thick walls and more or less circular lumina. Sometimes 



44 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



these are arranged into parallel rows as is normally the case, but in other leaves 
they may be quite irregularly arranged. All intermediates are to be found, but in 
some cases the margin provides valuable systematic characters. The two Com- 
bretaceous mangrove genera, Lumnitzera and Laguncularia, for example, are closely 
similar in cuticular characters, but one of the three main differences lies in the leaf 
margin. In Laguncularia (Plate 3 d) the cells are much smaller than those elsewhere, 
have thick walls and circular lumina and are arranged completely at random. In 
Lumnitzera, however, the marginal cells are scarcely smaller than the other epidermal 
cells, have little-thicker walls, have angular lumina, are rectangular and shghtly 
to conspicuously longer than broad, and are always arranged end to end in parallel 



rows. 



stomatal aperture 



outer stomatal ledge 




T-piece at 
stomatal pole 



poral \ walls of 

. , , f guard-cells 
epidermal J 



radial ) walls of 

> subsidiary 
tangential j cells 



Fig. 6. Terminology of stoma as seen in surface view. 

(4) Stomata. 

The term stoma is here taken to include the pair of guard-cells with the stomatal 
aperture or pore between them. In all vascular plants there are normally two guard- 
cells to each stoma, these arising from a single stomatal mother-cell at an early 
stage of leaf development, although abnormalities with one or three or more guard-cells 
do rarely occur (Dehnel, 1961; Ahmad, 1964a; Pant, 1965) and may even be of 
diagnostic value. Other cases have been described of bistratose guard-cells which 
have divided by a periclinal wall (Rehfous, 1914). The guard-cells are surrounded 
by normal or modified epidermal cells, which in the latter case are termed subsidiary 
cells or accessory cells. The whole system is known as the stomatal complex or 
stomatal apparatus. In surface view the guard-cells, in the great majority of 
vascular plants, and in almost all dicotyledons, are vaguely reniform. The anticlinal 
wall on the side towards the pore is here described as the poral wall, and that towards 
the rest of the epidermis the epidermal wall. If subsidiary cells are present their 
anticlinal walls abutting on the guard-cells may be known as radial walls, and those 
adjacent to the rest of the epidermis as tangential walls (Fig. 6). 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



45 



In vertical transverse section the guard-cells show an extremely various construc- 
tion which is surprising in view of their relative constancy in surface view (Fig. 7). 
The guard-cells vary in three main directions : their position relative to the 
subsidiary or other adjacent cells ; the relative thickenings of the various walls of 
the guard-cells ; and the pattern of cuticular ornamentation of the guard-cells. 
All these are important taxonomic characters. In most mesomorphic plants the 
guard-cells are on a level with or very slightly sunken below or raised above the rest 
of the epidermis, but in some xeromorphic plants they are often sunken to a 
considerable degree. In many conifers, as well as in a number of angiosperms, the 
guard-cells are very deeply sunken and completely overarched by the subsidiary 



back cavity 



sub-stomatal 
cavity 



stomatal fl 



nner stomatal ledge 



poral wall 
of guard-ce 




inner 



epidermal 



walls of 
'guard-cell 



subsidiary cell 

front cavity 



outer wall of guard-eel 
outer stomatal ledge 



stomatal aperture 
Fig. 7. Terminology of stoma as seen in transverse vertical section. 

cells, so that in a surface view the former are almost invisible, their position being 
marked by a ring of subsidiary cells (known as a poral ring) around a nearly circular 
pore. In other plants the subsidiary cells are also sunken. The thickenings of the 
guard-cell walls are very uneven in distribution, being connected with the physiology 
of the stomatal movements. In general, however, the poral and epidermal walls 
are rarely thickened, the former never, whilst the other two walls (i.e. the outer 
wall, adjacent to the environment, and the inner wall, adjacent to the inner regions 
of the leaf) are variously thickened. In some species, mostly xeromorphic, the 
thickenings are very considerable, and the cell lumen may be reduced to an extremely 



46 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

small proportion of the cell volume. The amount of cuticular ornamentation of 
the guard-cells is frequently proportional to the thickening of the cell walls, and 
hence to the degree of xeromorphism. The cuticular membrane which covers the 
epidermis is continuous over the subsidiary and guard-cells, and either ends where 
the inner and epidermal walls meet or also covers a varying amount of the 
subepidermal cells which line the space which usually exists below the stoma (the 
suh-stomatal cavity) . In most cases the cuticular membrane is very thin on the poral 
walls, but thick on the outer and inner walls, especially where each of these meets the 
poral wall. Here the cuticular membrane may be protracted into a pair of cuticular 
or stomatal ledges, ridges, horns, rims or hooks, which may be termed outer and inner 
respectively. This is found in many xeromorphic plants, but more frequently only 
the outer stomatal ledge is present. When the stoma is closed the poral walls of 
the two guard-cells meet, and above and below this point the outer and inner 
stomatal ledges also, if present, meet, or nearly so. Thus two extensions of the 
stomatal pore are delimited : 3. front cavity and a back cavity (Fig. 7) . In some plants, 
e.g. Bruguiera spp. (Areschoug, 1902) and Ceriops spp. (Stace, 1963), there are 
actually two pairs of outer stomatal ledges subdividing the front cavity into two 
portions. 

In cuticular preparations the stomata are usually well preserved, and they provide 
valuable systematic characters (cf . Plate 3 A, b) . Because much stomatal variation 
is in the third dimension (depth), however, a number of characters are not usually 
observable on cuticular preparations. It is often very difficult to ascertain the 
exact relative positions of the guard-cells and their neighbouring cells and the 
precise structure beneath the outer stomatal ledge. The cuticular membrane below 
the latter, since it curves back over the inner stomatal wall, is conspicuous on many 
preparations as a pair of stomatal flaps. A very good idea of the degree of cuticulariz- 
ation and sunkenness of the guard-cells and subsidiary cells may often be obtained 
by examining cuticular preparations with the membranes folded over on themselves 
inside outermost (Plate 5 c) . Also a surprising amount of structure can be observed 
once the many lines and ridges on the preparation have been interpreted by means of 
vertical sections. This approach was useful in the Rhizophoraceous mangroves, 
which have many diagnostic details of stomatal structure. The single known 
epidermal difference between Lumnitzera litter ea and L. racemosa, the shape of the 
guard-cells and subsidiary cells in section, however, was not visible from surface 
views of either cuticular preparations or whole epidermides. 

However, in surface view the most obvious cuticular characters are present on the 
outer walls of the guard-cells. The shape of the guard-cells as a whole is a useful 
criterion which is usually constant within species. The commonest shapes are 
circular, elliptic and oblong, but some stomata may even be broader than long, and 
sometimes angular. The stomatal poles where the two guard-cells meet may be 
obtuse, truncate, rounded or refuse, the last apparently being the case in many 
xeromorphic plants. The appearance of various thickenings on the outer guard-cell 
walls is also of importance. The outer stomatal ledge is usually visible from surface 
views, and its thickness and relative distance between the epidermal and poral walls 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 47 

are characters of significance. When this ledge is directly over the poral wall of the 
guard-cells it forms a thick rim to the stomatal aperture which Bandulska (1923, 
etc.) called the poral rim. At the stomatal poles there may be developed a T-shaped 
thickening or T-piece, much used by Bandulska (1924) as a character in the Fagaceae. 
Sometimes only the upright or the cross-piece of this is present. In some groups 
there is a small thin area at the polar end of each guard-cell, being circular, elliptic 
or comma-shaped. Solereder (1908) mentioned several examples and Rao (1939) 
found them commonly in the Magnoliales. A few genera are known to possess 
peculiar lobe-like extensions of the guard-cell epidermal walls, and these were 
noted in Sirephonema. 

Perhaps of more significance than any of the above characters, however, are the 
presence and types of subsidiary cells around the stomata. In his various studies 
Florin (1931, 1933) distinguished two types of stomatal development in the gymno- 
sperms, based on whether or not the guard-cells and subsidiary cells originate from 
the same mother-cell (syndetocheile and haplocheile respectively), and these types 
can usually be seen in the mature state because the former has two adjacent sub- 
sidiary cells and the latter a variable greater number. Although it was frequently 
necessary for Florin to draw conclusions concerning ontogeny by examining the 
mature anatomy, since many of his plants were fossils, it is unfortunate that it was 
so because the two terms are now more often used for describing the mature anatomy 
than the developmental sequence, and it has been shown by many workers that 
the presence of subsidiary cells does not necessarily signify a syndetocheile ontogeny, 
nor does their absence always indicate a haplocheile one. Thus the terms syndeto- 
cheile and haplocheile are better abandoned, and the cells neighbouring a stoma 
(whether they be subsidiary cells, i.e. different from the other epidermal cells, or not) 
described as mesogenous or perigenous according to whether or not they arise from 
the same mother cell as the stoma. These terms were first used by Florin (1933). 
In addition Pant (1965) proposed the term mesoperigenous for those situations where 
one neighbouring cell is mesogenous and the rest perigenous. This terminology has 
the advantages that it is equally applicable to all groups of plants rather than 
primarily to the gymnosperms, and that it allows for the fact that mesogenous 
subsidiary cells may become indistinguishable from other epidermal cells, or that 
perigenous cells may become specialized as subsidiary cells. The last term has no 
ontogenetic implications. 

The earliest system of subsidiary cell classification was adopted by Solereder (1908) 
from the work of Prantl (1872) and Vesque (1889). Six ontogenetic categories were 
recognized, three of which were perigenous and three mesogenous. Due to convergence 
during maturation only four types were recognizable in the mature state, being 
termed Ranunculaceous, Rubiaceous, Caryophyllaceous and Cruciferous. Metcalfe & 
Chalk (1950) completely ignored the modes of development and renamed these four 
categories as follows : anomocytic, with no distinctive subsidiary cells (Plate 4 a) ; 
paracytic, with 2 subsidiary cells lying parallel to the guard-cells (Plate 4 c) ; diacytic, 
with 2 subsidiary cells lying transversely to the guard-cells; and anisocytic, with 
3 subsidiary cells of which one is distinctly larger or smaller than the other two. 



48 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

These groups are often of considerable taxonomic significance, in many cases one 
type being characteristic of a whole genus or family. In the Comhretaceae, for 
instance, Strephonema is the only genus with paracytic stomata, and this agrees 
with the usual treatment of Strephonema as a separate subfamily. Apart from the 
four types of situation mentioned above there are other arrangements of subsidiary 
cells for which no terms are generally available, although for descriptive purposes 
they are obviously desirable. The term actinocytic was proposed by Metcalfe & Chalk 
(1950) for cases where the subsidiary cells, usually four or more in number, are 
elongated radially to each stoma, and cyclocytic by Stace (1963), where a similar 
number of cells forms a narrow ring round each stoma. Both terms have subsequently 
been used by other workers, although they probably refer to mere modifications of 
the anomocytic type. Of the 19 genera of Comhretaceae without paracytic subsidiary 
cells, in fact, two {Lumnitzera and Laguncularia) are cyclocytic (Plate 4 b) and the 
rest anomocytic. Bandulska (1931) also found cyclocytic subsidiary cells (" girdle 
cells ") in various genera of Myrtaceae. More usually, however, differentiation by 
subsidiary cell types is only at the specific level, or is even here absent. In Anopyxis 
klaineana (Pierre) Engler {A. ealaensis (De Wild.) Sprague), for example, paracytic, 
anisocytic and anomocytic stomata are said to occur on the same leaf (Metcalfe & 
Chalk, 1950), and other members of the Rhizophoraceae also show varying 
arrangements. It seems probable that only one basic mode of development is 
involved, but that the degree of subdivision of the subsidiary cells finally differs, 
producing the different arrangements. 

A reliable ontogenetic classification is highly desirable, and that proposed by 
Pant (1965) seems to be satisfactory. Ten categories are recognized, of which one 
is perigenous, three mesoperigenous, and six mesogenous. This nomenclature is 
quite separate from one based upon mature topography. The latter scheme would 
involve a larger number of categories (perhaps about 15), several of which could 
be derived from several different modes of development. Apart from the six terms 
mentioned previously one other has so far been named: tetracytic (Metcalfe, 1963), 
for cases where there are four subsidiary cells, two polar and two lateral, as in many 
monocotyledons. Terms such as paracytic, tetracytic, etc., could conveniently be 
prefixed by meso-, peri-, etc., once the mode of development was known. Members 
of the Ruhiaceae are meso-paracytic (Pant, 1965), for instance, and of the Gramineae 
peri-paracytic (Stebbins & Jain, i960; Stebbins & Shah, i960; Stebbins & Khush, 
1961). All investigated monocotyledons, in fact, have been found to be perigenous. 
It seems clear that the developmental studies at present being carried out mainly 
by Pant and his compatriots (e.g. Pant, 1965; Pant & Mehra, 1963; Pant & Verma, 
1963) will provide valuable evidence for a classification of stomata which will prove 
to be of considerable taxonomic and phylogenetic value. 

In some plants the subsidiary cells are furnished with distinct papillae, which may 
be absent elsewhere on the leaf. These may be quite long and overarch the stomatal 
aperture (Solereder, 1908). 

Stomata vary considerably in size and this is often a character of some importance. 
The relation between the level of ploidy and stomatal size is well known, but stomatal 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 49 

size does not always indicate the former (Carolin, 1954) . In many species the stoma- 
tal size also varies on a single leaf, but this may not prevent it from being an 
important taxonomic feature. On many leaves abnormally large stomata are to be 
found, and these are often, if not almost always, water-stomata, which are supposed 
to secrete drops of liquid water. Their presence is noteworthy. Water-stomata, or 
water-pores if the surrounding cells are not obviously guard-cells, may also be smaller 
than the normal stomata, and sometimes occur in groups (Solereder, igo8). 

The distribution and frequency of stomata are their most conspicuous characters, 
and are often of considerable systematic value. They are very often, of course, 
connected with the ecology of the species, and thus not likely to be of great phylo- 
genetic importance. In many hydrophytes, for example, stomata are either absent 
or infrequent and vestigial. In some terrestrial plants also some stomata may be 
vestigial, and Solereder (1908) describes situations where they may be plugged with 
a resinous mass. In plants with floating leaves they are usually on the upper 
epidermis only, and the same is true of other species with variously modified leaves 
which for some reason have a more concealed lower than upper epidermis, or even 
with apparently normal leaves (Solereder, 1908). In most dicotyledons, however, 
the stomata are more or less confined to the lower epidermis, although it appears 
to be usual to find a few scattered along the vicinity of the upper epidermal midrib 
of mesomorphic leaves. Solereder also noted this, and wisely pointed out that it 
could have " no great systematic value ". They are present there in most species 
of Comhretaceae, and are frequently abnormally elongated. They are sometimes 
present upon the midrib itself, rather than beside it, and cannot here be functional. 
All intermediate situations from the complete absence of stomata on the upper 
epidermis to their presence as abundantly as on the lower epidermis are to be found, 
even within the Comhretaceae alone. Lumnitzera spp., in fact, have stomata more 
abundant on the upper than on the lower epidermis. Species showing a slight 
increase from normal in the abundance on the upper epidermis have stomata 
scattered quite frequently alongside the midrib, more sparsely alongside the veins, 
and very sparsely in the rest of the areolae, and the further from the midrib they are 
situated the less abnormal is their anatomy. On the lower epidermis stomata are 
usually confined to the areolae, or non-venous areas, but sometimes they are to be 
found scattered on the midrib. They have seldom been recorded from the leaf 
margin of a dorsiventral leaf, or from the venules and veins of the lower epidermis. 

Thus stomata are typically found in the areolae of the lower epidermis, and 
usually only very sparsely anywhere else. In leaves showing various xeromorphic 
characters, especially where a closed system of venules is not apparent on the 
epidermides, they may be otherwise distributed. In the latter case they are frequently 
found all over the epidermis, except for the veins and midrib, but in some species 
they may be present only in special areas, e.g. longitudinal grooves, or depressions. 
These latter situations are not found in the Comhretaceae, although here there are 
some species with very prominent reticulations and thus sunken areolae. The 
position and types of sunken areas, the stomatal crypts or stomatal pits, may be very 



50 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

useful in taxonomy (cf. Morley, 1953). They are quite independent of the individu- 
ally sunken stomata found in a number of xeromorphic leaves. 

In leaves with stomata confined to areolae or to stomatal crypts the stomata are 
not regularly orientated, their long axes being quite randomly directed. In some 
leaves, however, especially narrow leaves or those with stomata present in longi- 
tudinal grooves, the stomata are all orientated parallel to the leaf axis. In a still 
smaller group of species the stomata are orientated at right-angles to the leaf midrib, 
e.g. Laguncularia racemosa (L.) Gaertn. f. These characters are usually of absolute 
constancy and of sufficient scarcity to be valuable in identification. 

The frequency of stomata, either in the areolae or in the non-venous areas, is a 
much used character, and frequently misused. A large number of workers have 
used the stomatal frequency in order to distinguish between species, with varying 
degrees of success. Timmerman (1927), for example, found the character of little 
use in Datura since the frequency varied considerably on different portions of the 
leaf. Environmental and other factors also cause variation in frequency. Valuable 
taxonomic evidence, however, may often be obtained if the stomatal index (looS/ 
E+S), based on the proportion of stomata to epidermal cells, is used instead of the 
frequency per unit area (Salisbury, 1927), as has been shown by various workers in 
modern and fossil plant taxonomy. In Lumnitzera, however, the two species only 
differ on stomatal frequency since the stomatal indices of the two species are 
identical, the stomatal frequency being greater in one due to the smaller size of aU 
the cells. 

(5) Types of trichomes and trichome-bases. 

Despite some attempts by Odell (1932) to prove otherwise, trichome anatomy 
provides a very important group of taxonomic characters. The definition of trichomes 
as outgrowths of the epidermis is not absolute as there is every grade between these 
and the so-called emergences which differ in also involving some subepidermal 
layers in the outgrowth, but in the present context both types are included. 

The extent to which trichomes are cuticular characters depends upon the degree 
to which they are cuticularized. In the case of most emergences and of trichomes 
with multiseriate basal parts a high proportion of the outgrowth is preserved on 
cuticular preparations, and the structure can be observed quite easily. Thus the 
whole or most of such outgrowths is cuticularized, only the more distal regions being 
sometimes not so covered. In the case of trichomes with pauci- or uni-seriate basal 
parts a similar proportion of the outgrowth may be cuticularized, but quite frequently 
only the very basal parts or even none at all possess a cuticular membrane. There 
are, of course, many exceptions to these generalizations ; for example the very long 
unicellular hairs of Macropteranthes leichhardtii F. Muell. are well cuticularized along 
their whole length. Where cuticularization is extensive cuticular preparations 
may provide the best means of studying the trichome anatomy. 

Outgrowths with large multiseriate bases are not readily broken off at the base, 
but when this happens an irregular hole appears in the cuticular preparation. 
Trichomes with pauci- or uni-seriate bases are much more easily removed in toto 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



51 



from the epidermis, the appearance of the remaining trichome-hase (i.e. the adjacent 
modified region of the epidermis, exclusive of any parts of the trichome itself) 
depending upon the original organization of the basal parts of the trichome, of which 
there are two main situations (Fig. 8) . Firstly the trichome may rest upon a normal 
or modified epidermal cell or group of cells, or, in the case of some unicellular 
trichomes such as root-hairs, the trichome may in fact not be separated from the 
epidermal cell by a cell wall. In these types the former position of a lost trichome 
will be indicated on the cuticular membrane by a modified epidermal cell or cells, 
by a hole in an epidermal cell, or will not be visible at all. Secondly the base of the 
trichome may be inserted between the epidermal cells, or may replace an epidermal 
cell, so that a lost trichome leaves a pore in the cuticular preparation. 

scarcely modified 
foot 




multicellular 
foot 



Fig. 8. Four types of trichome-base as seen in vertical section. 

Trichome-bases (hair-bases, gland-bases, etc.) of the above types are common 
features of cuticular preparations, especially as the trichomes are usually covered 
with a thinner cuticular membrane than the rest of the epidermis so that slight 
over-maceration removes the trichome completely. Their organization provides 
useful diagnostic characters, and it has been found necessary to create a terminology 
for the parts of the trichome-base (Fig. 9). The trichome-hase cells may or may not 
surround a pore. In the former case the trichome-base cell walls may be termed 
Poral, radial and epidermal, the first of these often being considerably thickened and 
then known as the poral rim. 



52 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



A trichome occupying the pore of a trichome-base therefore consists of two parts : 
the foot, which is inserted into the pore ; and the body, which extends above the 
epidermis. The foot may thus be a single cell, a small group of cells, or a peg-hke 
basal portion of a cell. The parts of the body of the trichome have been given very 
different names by various authors according to the trichome structure (e.g. stalk, 
head, branches, arms, disk, fringe, rim, etc.) and generalization is impossible and in 
act undesirable. 



base of poral rim 



hair-base cells 




poral rim 




epidermal) 

poral >walls of hair-base cells 

radial / 

Fig. 9. Terminology of two main types of trichome-base. 



scarcely 
modified foot-eel 



There have been many attempts to classify trichomes (cf. Bachmann, 1886 ; 
Solereder, 1908 ; Foster, 1949 ; Seithe, i960, 1962 ; etc.) but none has been 
generally successful. As mentioned previously, not only does each trichome type 
grade into the next but there is no real distinction between papillae, trichomes and 
emergences. The most usual division, between glandular and non-glandular types, 
suffers from the same disadvantages as other systems in that there is no sharp 
distinction between the groups, and in any case the division often separates obviously 
very closely related types. Furthermore, within each group or subgroup the direc- 
tions of variation are often quite different in various plant taxa, so that the only 
way to avoid the situation where a single trichome could belong to more than one 
group on different points of anatomy would be to create a system with many groups 
each with numerous parallel subgroups. It is thus quite clear that no useful general 
classification of trichomes can be constructed, but that instead it is advisable to 
create systems for each particular plant group investigated as it becomes necessary, 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 53 

and to employ special descriptive terms in each case without attempting to correlate 
the terms from taxon to taxon. This has the advantage that newly discovered 
types can easily be fitted into the system at any time. 

Good accounts of the ranges of variation and of the details of anatomy and func- 
tioning of trichomes are given by Solereder (1908) and Netolitzky (1932) respectively, 
and so no sort of review is given here. Instead an indication of the types of problems 
and the degree of taxonomic usefulness which may be involved in trichome studies 
is given by a brief statement of two major points of interest in the Comhretaceae. In 
all twenty genera of this family peculiar unicellular hairs with distinctive basal 
components (known as " Combretaceous hairs ") occur, elsewhere being known only 
in a few genera of the Myrtaceae and of the Cistaceae. These are of great interest 
for they virtually delimit the Comhretaceae from all other families except the other 
two mentioned above, and thus indicate a monophyletic origin of this family. The 
Myrtaceae, moreover, are generally considered to be closely related to the Comhreta- 
ceae, and the hairs thus support this view. The presence of the hairs in the Cistaceae, 
however, is difficult to understand since this family has never been considered at all 
closely related to the Comhretaceae. The identity of the Combretaceous and 
Cistaceous hairs thus needs to be investigated. In Strephonema, a Combretaceous 
genus usually placed in a separate subfamily from all the other genera, typical 
Combretaceous hairs are present in one of the species, thus confirming its inclusion 
in this family. In the other four species, however, they are absent and are replaced 
by hairs unique in the Comhretaceae, so that the hairs also support the treatment of 
Strephonema as a separate subfamily. 

The second point concerns the presence of " scales " (Plate 5 a), which are short- 
stalked trichomes with a disk-shaped multicellular head found in a wide range of 
plant families (Bachmann, 1886). These trichomes are often extremely distinctive, 
but in some families they intergrade with stalked globular glands, with stellate 
hairs, or with cup-shaped hairs. In the Comhretaceae scales are found in all species 
of the small genera Thiloa, Caly copter is and Guiera, and in over half of the large genus 
Comhretum. These four genera all belong to the same tribe [Comhreteae), and the 
scales do seem to indicate, therefore, that this classification is weU founded. In the 
Comhretaceae the disk of the scales is always one cell thick, but varies in the degree 
and manner of subdivision, giving rise to a very wide range of types. In Rhododend- 
ron, a genus of the very distant family Ericaceae, however, the main direction of 
variation is in the overall organization of the scale rather than in its division in one 
plane, producing an equally varied range of forms (Cowan, 1950 ; Seithe, i960). 
Furthermore, in Rhododendron there are intermediates between all of the numerous 
(25 according to Cowan ; 43 to Seithe) types of trichomes, whereas in the Comhreta- 
ceae there are three main types (Combretaceous hairs, scales and stalked glands) 
which are always quite distinct (Stace, 1965). Thus Cowan concluded that aU the 
trichomes of Rhododendron are phylogenetically directly related, especially as they 
all appear to pass through identical early stages of ontogeny. Although this may 
be correct in Rhododendron I do not consider it to be the case in Comhretum, where it 
is not possible to construct a " phylogenetic tree " of trichome types as did Cowan. 



54 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

The stalked glands of the Comhretaceae (Plate 5 b) are found in all the scale-less 
genera of Comhreteae except Pteleopsis, in all the scale-less species of Combretum 
itself, and also in Conocarpus, a genus in another tribe. Their presence in the last 
genus and absence from Pteleopsis is puzzling, since the two tribes are otherwise 
well defined on a number of characters. The presence and types of stalked glands 
or scales are thus very useful taxonomic criteria in Combretum and in other genera 
of the Combreteae (Stace, 1961, 1963, 1965 ; Exell, 1963) . In both Combretum and Rhodo- 
dendron the major taxonomic positions of trichome differentiation are at the sub- 
generic and sectional level, and where a section contains species with unexpectedly 
divergent trichome types there is good reason to doubt the homogeneity of that group. 

Compared with trichome structure trichome size is relatively unimportant, but 
nevertheless at times very useful. For example the Combretaceous hairs of the 
Combretaceae are relatively very uniform in structure, but those of Macropteranthes 
leichhardtii are very distinct in their large size, and those of Outer a senegalensis 
extremely so in their very small size. The length of hairs generally is, of course, a 
frequently used taxonomic character. 

The trichome-base frequently provides characters of importance, most notable of 
which is the shape (i.e. the degree of modification) of the trichome-base cells. In 
some cases they are scarcely modified from normal epidermal cells, but usually they 
are elongated radially to the pore, when this is present. If the normal epidermal 
cells are undulate-walled the radial walls of the trichome-base cells are very often 
straighter or straight, and frequently their epidermal walls also. Although the 
number of trichome-base cells varies considerably within one species, in many cases 
there is a typical range, of which 5-8 is the commonest, which may aid in the 
diagnosis of the plant. Some species, for example, have a characteristically low or 
high number of trichome-base cells. In some species the trichomes are so numerous 
that almost every epidermal ceU is a trichome-base ceU, and radially elongated. In 
Guiera (Plate 4 d) many upper epidermal cells are adjacent to two or three hair-bases, 
and there are no unmodified upper epidermal cells. The latter are thus typically 
elongated in two or three directions, and the cuticular preparations of this genus 
are quite distinct in the whole family. In many Combretum species (e.g. C. glutinosum 
Perrott. ex DC. and several related species) almost all of the lower epidermal cells 
are scale-base or hair-base cells. 

Quite frequently, especially in xeromorphic species, the trichome-base cells are 
not radially elongated, but have become tangentially divided so that there are 
several rings of modified cells, the inner of which is composed of cells longer in the 
tangential than in the radial plane. The cuticular flanges of these walls are usually 
thick and straight. Sometimes the two different types of trichome-bases mentioned 
above form valuable diagnostic characters between species, although in other cases 
this is not so and too much emphasis should not be placed on them. In Terminalia 
glabrescens Mart., for example, both extremes and all degrees of intermediates may 
be found. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 55 

The size of the pore, where present, varies considerably, but in some cases may 
characterize species. In Anogeissus acuminatus (Roxb. ex DC.) Wall, ex Bedd., for 
example, the pores are very large (often over 50 [jl across) and are a quite distinctive 
character of this species. The thickness of the poral rim varies in two dimensions, 
i.e. towards the centre of the pore (Plate 5 a), and towards the centre of the leaf. In 
the former case it may result in a very small pore of only 1-2 [x in diameter. The thick 
rims are frequently, though not always, accompanied by a considerable thickening 
of the radial walls of the trichome-base cells, so that a stellate pattern of thickening 
is obtained. In extreme cases, where the trichome-base cells are of the tangentially 
subdivided type, all the walls of the trichome-base are conspicuously thicker than 
the rest of the epidermal cell walls. The poral rim often becomes deeper by a 
development of cuticular membiane around the foot of the trichome, often up to the 
inner wall of the epidermis. The diameter of the pore is nearly always narrower at 
the base than at the apex, so that the sides and base of the pore can be seen in 
surface view. 

Where there is more than one type of trichome on a leaf the types of trichome-bases 
bearing them are often different (Plate 5 d). In Combretum, for example, where 
hairs are always accompanied by either stalked glands or scales, the poral rims of the 
latter are much thicker and deeper than those of the hair-bases, and the two can be 
readily distinguished. Moreover the distribution of the scales is different from that 
of the stalked glands, the latter being commoner on the veins, the former in the 
areolae. Thus the abundance and distribution of each type of trichome can be 
ascertained even if the trichomes have been lost. In Conocarpus erectus there are two 
types of bases : those with radially and those with tangentially elongated cells. 
Stalked glands are found on all of the latter types, on intermediate types and on a 
few of the former types ; whilst Combretaceous hairs are confined to the former 
types of bases. Thus in this genus the difference is not so well marked. 

(6) The cuticle. 

Besides providing a means of studying epidermal anatomy in the absence of any 
cellular layers the cuticle often possesses characters of its own on its outer surface, 
and the cuticular membrane also varies in the degree of its development. All these 
characters are worthy of study as potential taxonomic criteria. 

The most obvious feature is the degree of development of the cuticular membrane, 
as has been mentioned in several preceding sections, notably the cuticular orna- 
mentation of the guard-cells and trichome-bases. The thickness of the cuticular 
membrane over the normal epidermal cells is also a character of diagnostic importance. 
In most cases thicker membranes are characteristic of plants of drier habitats, and 
are thus a measure of xeromorphy, although this is not always so. In Lumnitzera 
racemosa and Macropteranthes kekwickii F. Muell., for instance, the wall between the 
lumen of the epidermal cells and the exterior is quite thick, as might be expected in 
these two xeromorphs. However, in the former the cuticular membrane is thick 
and has long stout cuticular flanges, whilst in the latter the cuticular membrane is 



56 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

thin with scarcely any flanges, the non-cutinized epidermal wall accounting for the 
greater part of the thickness of the outer compound wall. Because of its relation 
to the environment, however, the degree of development of the cuticular membrane 
is liable to considerable phenotypic variation, and great care must be exercised in 
its use in taxonomy. The cuticular flanges show the same degrees of genetic and 
phenotypic variation and may also prove of diagnostic value in some cases, besides 
determining the over-all appearance of the cuticular preparation and whether or not 
the epidermal cell outlines are visible (Plate 5 c). In Combretum psidioides, for 
example, the cellular organization of the scales is always more or less completely 
indiscernible, whilst in all other closely related groups it is clearly visible on cuticular 
preparations. Again, xeromorphic species usually have well-developed flanges, but 
some species have thick cuticular membranes yet very short and narrow flanges. 

The other character of the cuticular membrane is the type of minute ornamentation 
of the outer surface of the cuticle, this having been used by a number of past workers. 
The vast majority of plants have a smooth or minutely granulated cuticle when 
viewed with a light microscope, and no use can be made of this fact beyond the 
absence of any definite markings. The latter, as seen in many mesomorphic plants, 
take the form of a series of minute grooves and ridges which are commonly termed 
striations from their appearance in surface view. Their presence and degree of 
development do not always seem to be related to xeromorphy, although some 
species of Comhretaceae have been found in which the degree of striation is not 
constant. In their less-developed state the striations are usually found running 
parallel to the midrib and veins, and radially to the trichome-bases and stomata, and 
the presence of such a pattern is probably of little importance. In general the 
striations bear no relation to the cell outlines, each marking passing over a number 
of cells, but exceptions have been found where each striation is always contained 
within a single cell outline. When the striations are better developed they appear 
on the non-venous areas unconnected with the trichomes or stomata, and their 
orientation, arrangement and degree of development may be very various. Where 
they are found in areolae they are usually contained within one areola, i.e. they do 
not continue across the venules. Where the veins are less extensive on the epidermis 
the striations may be much longer. They are extremely conspicuous, for example, 
in Macropteranthes kekwickii and M. montana (F. Muell.) F. Muell., where they are 
found on all parts of the epidermis. In section the surface of the cuticle appears 
papillate. In extreme cases, several of which have been found in Combretum, the 
whole cuticle is covered with striations, those radial to the trichomes and stomata 
and those parallel to the veins being continuous over other areas of the epidermis 
(Plate 5 d). Sometimes the striations are so strong, and the cuticular flanges so 
weak, that the former obscure all signs of the cell outlines. 

In Combretum zenkeri the long papillae have a conspicuously echinate cuticle. 
The series of raised cuticular anastomosing ridges reported by Solereder (1908) in 
several families appear to be rare. 



cuticular studies as an aid to plant taxonomy 57 

§6. Description of cuticular characters 

Just as the formal description of a new plant species follows a well-defined, and 
well-known pattern, so should the characterization of a cuticular membrane, although 
hitherto this has not been the case. Two types of description are required for various 
purposes : a relatively full account of the cuticular membrane characters ; and an 
abbreviated account or diagnosis. A suggested form for the former, after many 
adjustments, is as follows : 

1. Distribution of species ; habitat. 

2. Morphology of the leaf : 

a. general morphology as seen with the naked eye. 

b. special morphology — venation pattern and prominence, trichome types 
and distribution, clear spots, pimples, etc. 

3. Account of the cuticular membrane : 

a. general features — thickness, ease and quality of preparation. 

b. shape, arrangement and size of upper epidermal cells of : 
i. venation system ; 

ii. non-venous areas. 

c. shape, arrangement and size of lower epidermal cells of : 
i. venation system ; 

ii. non- venous areas. 

d. shape, arrangement and size of margin cells. 

e. distribution, frequency, orientation, size and structure of stomata, in- 
cluding the subsidiary cells. 

f. distribution, frequency, size and structure of trichomes and trichome- 
bases. 

g. presence of any other features, e.g. cork-warts, hydathodes, water-stomata, 
domatia. 

h. presence of markings, e.g. striations, on the cuticle. 

4. If necessary : interpretation of surface cuticular characters, primarily of the 
stomata, by means of leaf sections. A brief mention of the organization of the 
hypodermal and inner layers of the leaf might not be out of place as a general 
background to the cuticular characters. 

Such a description is an obvious prerequisite for the use of cuticular characters in 
taxonomy and identification, and full allowance must be made for environmental and 
other variation. Characters which are thought to be hkely to vary beyond the 
limits observed should be indicated. 

For the purposes of a diagnosis the third section above, i.e. the description of the 
cuticular characters themselves, should alone be used, and this in an abbreviated 
though essentially similar form to that above. To this end certain abbreviations 
may be employed, comparable to Ivs. , fls. and fr. in other descriptions. The following 
are suggested : epid. (epiderm-is, -ides or -al) ; 1. and u. (lower and upper) ; cutic. 
(cuticle (s), cuticular or cuticular membrane (s), unless these need to be distinguished) ; 



58 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

stom. (stom-a, -ata or -atal) ; ven. (venule (s)) ; ret. (reticulum) ; w.l./wall (wave- 
lengths per wall) ; ampl. (amplitude) ; isod. (isodiametric) ; 2ry (secondary) ; lat. 
(lateral (ly)) ; longit. (longitudinal (ly)) ; trans, (transverse (ly)) ; tang, (tangential 
(ly) ; rad. (radial(ly)). 

In some cases these procedures may need to be modified, just as with normal 
taxonomic descriptions, but in general their use would lead to a greater uniformity 
and thus ease of reference and comparison. A diagnosis of a typically meso- 
morphic species is given here : 

Diagnosis of Combretum fruticosum (Loefl.) Stuntz. 

Cutic. medium thickness ; features clear. U. epid. : midrib broad and conspicuous, of many 
longit. rows of narrow elongated cells ; lat. veins conspicuous but narrower ; minor veins and 
some ven. distinguishable but ven. ret. absent or fragmentary ; cells of non-venous areas 
averaging c. 30 pi across, isod., with straight walls, or less often with undulate walls with up to 
i^ w.l./wall and 5 jjl ampl. L. epid. : midrib and lat. veins broader and more conspicuous than 
on u. epid. but minor veins and ven. not so much more conspicuous and ven. ret. still ill-defined ; 
cells of areolae averaging c. 20-25 y- across, isod., with faint undulate walls with up to 2^ w.l./wall 
and 8 (j, ampl. Margin of a very few rows of slightly elongated rectangular, straight-walled 
cells. Stom. very frequent in non-venous areas of 1. epid., sparse alongside u. epid. midrib, 
randomly orientated on 1. epid., c. 22-30 x 15 (x ; epid. walls very faint ; poral walls quite 
conspicuous ; stom. ledges absent ; poles mostly rounded, sometimes retuse or obtuse ; 
adjacent epid. cells 4-7, unmodified. Combretaceous hairs frequent on midribs and lat. veins 
of both epid., but sparse elsewhere ; hair-bases with pore c. 10-25 i'- across ; poral rim some- 
what thickened ; hair-base cells mostly 5-9 and slightly rad. elongated, with straight, slightly 
thickened rad. walls ; internal compartments long and pointed. Scales bowl-shaped with 
raised convex cutic. membrane, frequent in 1. epid. areolae, sparse on 1. epid. venous areas and 
on u. epid., there frequently commoner on or near midrib ; scale-bases with pore c. 10-25 \^ 
across ; poral rim considerably thickened ; scale-base cells c. 8-13, considerably rad. elongated 
with thin straight rad. walls ; scales c. 100-200 (x across, circular in surface view, scalloped at 
margin, divided into c. 35-70 cells by rad. walls alone, few of the cells reaching the scale centre, 
all reaching the scale margin ; stalk of scale uniseriate. 

§7. Variation of cuticular characters 

The main disadvantage of many taxonomic characters is that they will vary 
independently of the genotype, and it is thus important to discover to what extent 
and under which conditions this variation will occur. In all groups certain characters 
vary so much that they are useless in taxonomy, but since these characters are not 
the same ones in every plant group every character is worthy of study. With regard 
to cuticular characters variation is due to three major causes : the age of the leaf, 
or rather its degree of maturity ; the environment in which the plant is situated ; 
and the position of the leaf upon the plant, that is to say the internal and external 
environment of the leaf. In general the characters affected by these three factors 
are those of size, frequency and degree rather than the actual anatomy or organi- 
zation of the particular structures. Thus the majority of the characters previously 
discussed show relatively little of this type of variation. Much of the variation 
shown within species is of course genetically controlled ; this must be equally 
carefully studied, and, if possible, distinguished from the environmental and develop- 
mental variation. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 59 

In the case of variation due to age it is obvious that in extremely young leaves 
every cuticular character would be quite unlike that character on a mature leaf, and 
such immature leaves are usually of no taxonomic use with regard to either macro- 
scopic or microscopic features. The present section therefore only deals with leaves 
that are likely to be chosen for cuticular examination, i.e. leaves which are over 
half-grown. 

The relevant literature shows a generalized sequence of the development of the 
various cellular layers of a typical simple leaf to be established (Avery, 1933 ; Esau, 
1953 ; Slade, 1957 ; etc.). The leaf primordium arises as a bud-like outgrowth of 
the apical meristem of the stem, and develops by its own apical meristem into a 
finger-like projection which corresponds to the petiole and midrib. The lamina 
later grows out from this as a pair of lateral wings, this activity of a marginal meri- 
stem usually commencing before the cessation of that of the apical meristem, in 
most cases before the leaf is a millimetre long. When apical growth has ceased the 
leaf elongates by intercalary growth. The thickness of the lamina in terms of cell 
number is determined at a very early stage in development, when the leaf is a few 
millimetres long, and subsequent increase in cell number is solely by anticlinal 
divisions. Very soon after the lamina has begun to develop the major lateral veins 
appear in its central portion, branching from the midrib and keeping pace with the 
lamina extension bj^ intercalary growth. The lesser veins and venules develop 
similarly at progressively later stages. Recent work (Pray, 1963) has shown that 
the veinlet terminations also develop in the same manner, and not by dissociation 
of parts of the venule reticulum in the expanding leaf as was formerly thought to 
be the case. 

Before the leaf approaches its mature size cell division ceases altogether, the 
remaining increase in leaf size being by cell expansion alone, and, although some of 
the layers are distinguishable before cell division ceases, most of them acquire their 
characteristic form after this point. The first parts of the leaf to end cell division 
are the two epidermides, followed by the spongy mesophyll and lastly the palisade 
mesophyll. Not all regions of the leaf stop producing more cells at the same time, 
moreover, the leaf apex reaching maturity first and the base last, and superimposed 
on this sequence the leaf margin continues cell division longer than the midrib region. 
The tip of the leaf is thus the most mature, and the marginal parts the least. Since 
the epidermis is the first layer to show a cessation of cell division, cell enlargement 
shows its greatest development there. Stomata develop from epidermal initials 
soon after the number of cells in the lamina thickness has been established, and their 
mature shape is attained in most cases before that of the other epidermal cells. 
The times and rates of trichome development usually appear to run parallel to those 
of the stomata, and in almost all the cases investigated trichomes and stomata had 
reached their mature structure before the leaf was its mature size. Thus one of the 
latest features of the epidermis to appear is the mature epidermal cell shape and 
size. 

Examination of data from a wide range of taxa has led to the conclusion that there 
are six cuticular characters which are regularly to be expected to show sufficient 



6o CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

environmental and developmental variation to pose serious taxonomic difficulties. 
These are discussed separately. 

A. Epidermal cell size. 

A number of workers (e.g. Yapp, 1912 ; Salisbury, 1927 ; Watson, 1942) have 
found that the epidermal cells are larger on leaves in more humid or more shaded 
situations, and Odell (1932) cites workers who also discovered a decreasing cell size 
with excess carbon dioxide, drier air, drier soil and greater altitude. In general 
these size differences are proportional to a change in leaf area. The cell size also 
decreases with an increased height of insertion upon the axis of many herbs (Yapp, 
1912 ; Stober, 1917 ; Odell, 1932 ; Turrell, 1942 ; Ashby & Wangermann, 1950 ; 
etc.). Although this might seem to reflect the more shaded and humid environment 
of lower leaves Ashby & Wangermann (1950) concluded that the difference in cell 
size was not dependent upon the climate when the leaf unfolds, even though in 
ipomoea at least the later the time of seed sowing the larger were the epidermal cells, 
but found (1950a) that the factors affecting this character operate mainly by in- 
creasing the period of cell division in upper leaves. Ashby (1948) discovered that 
the decrease in cell size was accompanied by a decrease in leaf area and cell number, 
and although the latter were affected by water supply the former was not, since 
plants in dry and wet conditions produced the same-sized epidermal cells. This is 
contrary to the results of most of the above workers. 

Some investigations into the relation between cell size and leaf maturity were 
made by me on a number of species, using North's cellulose acetate film technique 
(North, 1956). In the two main species investigated, Quercus robur and Carpinus 
betulus L., the leaves grow by cell division and enlargement until they are about 
one-third their mature size, when cell division ceases and subsequent leaf enlarge- 
ment is entirely due to cell enlargement. In species with undulate epidermal ceU 
walls, e.g. C. betulus, the cells appear for a time to be enlarging more rapidly than 
the leaves. This apparent anomaly is due to the fact that cells can increase in 
length and breadth without any increase in area while the wall undulations are 
developing (cf. Fig. 10). 

Secondly, the nature of the abnormally small leaves found at the base of most 
sucker shoots was investigated in Quercus robur, Buchenavia capitata (Vahl) Eichl. 
and Platanus X hybrida Brot., cells from fully mature leaves of as wide a size range 
on the same sucker as possible being examined. No significant cell-size difference 
whatsoever was found, an increase in mature leaf size of a factor of three producing 
no increase at all in the cell size in all three species. Representative figures obtained 
from Quercus and Buchenavia are : 

Quercus Buchenavia 

Lf. size Av. cell Igth. Lf. size Av. cell Igth. 

9-0 X5-0 cm. 30-0 (jt. 3-5 X2-0 cm. 42-5 [x 

6-0 X3-0 cm. 28-0 [X 2-5 xi-3 cm. 4i'5 H- 

4"5 X3-0 cm. 31-0 ]x i-o xo-6 cm. 40-5 [x 

3'3 X2-0 cm. 28-5 [X 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 6i 

This is in direct contrast to the results obtained by Turrell (1942) and others working 
with herbs showing typical heteroblastic development (i.e. varying leaf types up the 
stem) where the upper and lower leaves differed in a number of size characters 
besides the overall leaf size, and thus indicates that the small leaves at the base of 
suckers do not represent normal heteroblastic growth. Since different-sized leaves 
on a single sucker differ in cell number but not cell size the factors regulating the 
mature leaf size must take effect early on by determining the point at which cell 
division stops. The same results were obtained from abnormal suckers on coppiced 
trees of Quercus which possessed leaves over twice as long and twice as broad as 
those on normal suckers, yet which had epidermal cells of the same size. 

Finally, the average epidermal cell sizes of different parts of a single leaf were 
measured, when in every case the cells increased in size with the distance from the 
leaf apex and from the leaf margin. Representative measurements for Quercus 
robur are : extreme leaf apex, 29 X22 [x ; extreme lamina base, 36 X31 [x ; extreme 
margin, half-way from base to apex, 30 X20 (x ; half-way from base to apex and 
from margin to midrib, 34 X27 [x. 

These observations merely serve to emphasize the variability of epidermal cell 
size not only with age and minor genetic variation but with the environment, 
position of the leaf on the shoot and the position of the cells in the leaf. However, 
there does appear to be less variation than might be expected with respect to the 
cells of leaves on a single shoot which attain different sizes at maturity, and this is 
encouraging from the angle of systematic anatomy. Should the character appear 
to be of iaxonomic use great care must be taken to use strictly comparable mature 
material, and to ascertain the degree of variation to be found. Thus with herbarium 
material this feature should not be used unless very well marked. 

An additional compUcation, the secondarily subdivided epidermis, is unlikely to 
cause any confusion as the nature of the extra walls is usually obvious (see §2 above) . 
They presumably develop long after normal cell division has stopped, and represent 
a renewal of meristematic activity. 

A very similar situation to the above exists regarding stomatal size, although in 
this case the leaf size is not wholly and directly dependent upon the character. 
Measurements indicate, in fact, that the stomata may become mature in size 
and anatomy before the epidermal cells have fully expanded, and so the use of 
absolutely mature leaves may not be so essential. Various workers, especially 
those previously mentioned, have found that stomata become longer nearer the 
base of a herbaceous plant, in shade, in moist air and in moist soil. These results 
are not always capable of repetition, however, and some workers (e.g. Odell, 1932) 
found no difference in stomatal size on sun and shade leaves. Thus the variation 
of this character in the particular group concerned should be fully ascertained 
before it is taxonomically employed, but the evidence does suggest that the average 
stomatal size is less subject to phenotypic variation than that of the epidermal cells. 



62 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

B. Stomatal frequency. 

Since the number of stomata that appear on a leaf is determined at the end of the 
period of cell division, long before a leaf reaches its full size, the stomatal frequency 
expressed as the number of stomata per unit area clearly decreases as the leaf 
expands. Again, therefore, only mature leaves are suitable for comparative 
measurements of stomatal frequency. Moreover the stomatal frequency often 
varies considerably on different parts of the same leaf, and on different leaves of the 
plant. Salisbury (1927) found that the lowermost leaves of a plant, and the basal 
and the midrib regions of the leaf, had a lower stomatal frequency, which was 
reflected to an equal extent on both epidermides when the upper epidermis possessed 
stomata. The increased frequency of stomata on higher (smaller) leaves of herbs has 
also been noted by Yapp (1912), Stober (1917), Rea (1921) and Turrell (1942), but 
Odell (1932) and others have found that the frequency increases towards the base 
and midrib of each leaf. 

Environmental factors also regulate the stomatal frequency, the latter being 
lowered by humid air, wet ground, shade from sun and wind, lower altitude, and a 
lesser concentration of carbon dioxide {fide Odell, 1932). 

Fortunately, however, the above variation can apparently be almost entirely 
cancelled by recording stomatal frequency in terms of the proportion of stomata to 
epidermal cells : 

No. of stomata , -r -. 

r:; 7-— r ^r? z — n \ — T7" Xioo =Stomatal Index. 

No. of stomata + No. of epidermal cells 

Since the variation in stomatal frequency caused by the above factors is apparently 
due to the increase or decrease in epidermal cell (and to a lesser extent stomatal) 
size, but not to cell number, the stomatal index is not affected. Salisbury (1927), 
who was the first to use this value, found that one factor did, however, alter the 
stomatal index — the humidity. It is not known how widespread this phenomenon 
is, but it seems that the diverse variation shown by stomatal frequency can often 
be more or less overcome. More work is needed, however, to establish that the 
stomatal index is always as invariable as has been claimed : Yapp's (1912) data, 
for example, would suggest that it is not. Gupta (1961) also used a value which 
cancelled the effect of the position and environment — the absolute stomatal number. 
This is obtained by calculating the product of the stomatal frequency per unit area 
and the leaf area, and can therefore be used as an alternative to the stomatal index. 
Reyenga & Karstens (1964) described an unusual situation in the sepals of 
Hydrangea, which are small and pink during anthesis, but which afterwards turn 
green and enlarge considerably. This period of growth includes not only the 
formation of new stomata but also the disintegration and disappearance of some of 
the original ones, the stomatal index rising due to the predominance of the former 
process. Moreover, the original stomata have actinocytic subsidiary cells with 
radial cuticular striations, whilst those of the later stomata are anomocytic and 
not striated. Furthermore, Dehnel (i960) reported in Begonia the disintegration of 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 63 

stomata which have been wounded. Phenomena of this kind are probably- 
exceptional, however, and need rarely enter into consideration. 

Rea (1921) found that in Campanula rotundifoUa L. the hydathodes showed the 
same positional and environmental variation as the stomata. 

C. Trichome frequency. 

The general situation regarding trichome density is similar to that of stomatal 
frequency in that usually the number of organs is determined when the leaf is very 
young and the frequency thus drops as the leaf expands. Thus a densely pubescent 
young leaf may develop into a sparsely pubescent mature leaf. Yapp (1912), 
however, found that in Filipendula ulmaria (L.) Maxim, there was an additional 
development of hairs after the leaf began to unfold, the original pubescence being 
developed in the leaf bud. As in the case of stomata the pubescence often varies 
from leaf to leaf, from one part of a leaf to another, and under a variety of environ- 
mental conditions. The variation in trichome frequency on different parts of one 
leaf is a well-known and important taxonomic character. Yapp (1912) and Stober 
(1917) both found that the pubescence on plants generally rose with increasing 
distance up the aerial shoot : in many plants, in fact, the radical leaves are almost 
glabrous and the upper cauline leaves conspicuously pubescent. The inner radical 
leaves are also usually more pubescent than the outer. 

The same workers, in general, as those who have found a variation in other 
characters, have also documented the variation of trichomes in different environ- 
mental conditions {fide Odell, 1932). Hairs have been found to be more abundant 
in greater sunlight, greater wind exposure, drier air, drier soil and greater altitude. 
McDougall (1927), however, reported that far more hairs developed in reduced 
sunlight in Lactuca biennis (Moench) Fernald (L. spicata auct.). It is possible that 
in some of these cases the proportion of hairs to cells does not vary, but at least in 
many instances it does. In general there is no possibility, therefore, of the use of a 
trichome index comparable to the stomatal index ; trichome density is a much more 
variable character than stomatal frequency, and its taxonomic use is thus more 
restricted. Only differences of a considerable degree or differences in the relative 
distribution of trichomes are reliable. The one advantage of a measurement of hair 
density in terms of hairs per cell rather than of hairs per unit area is that the former 
cancels the effect of age in any one leaf, except as noted below. As in the case of 
stomata the number of hairs per epidermal ceU may remain constant as the leaf 
expands, and it is unfortunate that this value does not also remain constant in 
different environmental conditions and positions on the plant. 

A further complicating factor concerned with trichome frequency is the 
phenomenon of glabrescence : the gradual loss of hairs with age. As explained 
previously most hairs may be localized on cuticular preparations by the presence 
of a hair-base, the centre of this often being occupied by a pore. It was thought 
that the number of hair-bases per epidermal cell might remain constant throughout 
the hfe of a leaf, even if most of the hairs themselves dropped off. Unfortunately 



64 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



in most cases this was found to be untrue. Observations were made chiefly on a 
number of species of Buchenavia, Conocarpus and Comhretum ranging from 
extremely small young leaves, from which cuticular membranes were difficult to 
prepare, to senescent leaves that had begun to fall from the trees. In the species of 
Buchenavia examined, notably B. capitata and B. kleinii, the very young leaves are 
uniformly densely pubescent or sericeous, whilst the very old leaves are mostly 
glabrous with a few hairs still remaining on the midribs and major lateral veins. 
Microscopic examination of the former showed that almost every epidermal cell 
was adjacent to one or more hair-bases, but as the leaf expanded and the hairs began 
to fall off the hair-bases gradually became occluded, and in the mature state the 
majority were not visible at all. The more actively growing (i.e. the more immature) 
a leaf the more quickly occluded are unoccupied hair-bases. Occlusion is effected 
by the hair-base cells encroaching inwards until no pore remains. When loss of 
the trichomes is delayed until after leaf growth has ceased, however, the vacated 
hair-bases often remain distinct thereafter, as is the case with the diagnostically 
important scales of Comhretum spp., although this is not true of the hairs of 
Conocarpus erectus var. sericeus. 

Thus the use of trichome frequency as a taxonomic character must be preceded 
by an exhaustive series of observations on a wide range of material in order to 
discover the cuticular changes accompanying growth and senescence, as well as the 
degrees of phenotypic variation. 

D. Epidermal cell shape. 

The epidermal cells gain their characteristic shape gradually from the time that 
cell division ceases onwards, and in general this is completed before the process of 
cell expansion has ended. Certain types of cell shape, however, may be finally 
assumed after the leaf has reached full size. The only important example of this is 
the undulation of the lateral epidermal cell walls. Since the undulations occur 
equally on either side of an imaginary mid-line, i.e. into each of the two adjacent 
cells, the apparent length of the cell can increase (by twice the value of half the 
amplitude) without any increase in volume (Fig. lo) . 

original length = I 
original breadth — b 
original area = lb 

original cell wall 

final cell wall 

amplitude = x 
final length = / +x 
final breadth — b + x 
final area = lb 

Fig. io. Diagram illustrating growth of cell-wall undulations. 

In most cases examined, however, the development of undulations takes place 
whilst the leaf is expanding, and can give rise to the apparent phenomenon that the 




i 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 65 

cells are growing faster than the leaf. The final amplitude of the undulations is 
usually reached as the cells (and the leaf) attain full size, although the maximum 
growth of the undulations ceases some time before this. In Buchenavia capitata, 
however, the leaves at flowering time, which are fully grown but considerably 
thinner and more pubescent than when mature, have conspicuously less undulate 
walls than those at fruiting time, and in this case at least the undulations do increase 
in amplitude without any increase in cell area. This is most apparent on the upper 
epidermis. According to Watson (1942) it is the stresses set up by the hardening 
cuticular membrane which cause the plastic cell walls to become undulate, and when 
the cells have become mature their walls are too rigid to allow further increase of 
undulation. It is thus clear that absolutely mature leaves must be examined in 
order to ascertain the degree of undulation of the cell walls. 

Much variation in this undulation has been noted in various environmental 
conditions, strongly undulate to straight walls often being observed within a single 
species. Brenner (1900), amongst the earlier workers, found that considerable 
differences could be found between plants in " normal " and damp conditions : in 
two species of Crassulaceae the normally straight-walled cells were converted to 
markedly undulate-walled cells, and in a third species the exactly opposite result was 
obtained. A considerable number of workers have noted that the amplitude of the 
undulations increases with increased shade, humidity and soil dampness (Odell, 
1932). Watson (1942) found that epidermal cells of Hedera in sunny situations had 
6-3 crests per cell, whilst those in shade had 8-7. Moreover the latter extended 
throughout the height of the cells, whereas the former were confined to their outer 
edge. This was explained on the basis that in the sun the cells become mature 
and rigid more quickly, so that the hardening cuticular membrane does not have an 
effect to such a great depth as in the shade. 

Yapp (1912), Stober (1917) and other workers {Jide Odell, 1932) have found that 
the undulation of the cells of many species decreases the higher the point of insertion 
on the stem, in some cases radical leaves having undulate-walled cells and cauline 
leaves straight-walled cells. 

With regard to other features of cell shape Brenner (1900) claimed that Crassula 
portulacea Lam. has strongly papillose cells in damp conditions, but flat cells in 
" normal " situations. 

Few other characters of the epidermis show as much variation as the cell-wall 
undulation, and although this has been used taxonomically in a number of groups 
its use requires the most extreme caution. 

E. Leaf venation. 

Variation in the actual venation of leaves is apparently not very frequent, and 
when it does occur not drastic. The characters found to vary, if any, are exemplified 
by the situation found in Medicago sativa L. by Turrell (1942), where the larger leaves, 
which had larger epidermal cells and stomata than normal, possessed larger areolae, 



66 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

i.e. the number of veins and venules was not different, but the spaces between them 
were larger. Levin (1929) and Hall & Melville (1951) claimed that the areola size 
was diagnostically important, but the latter authors (1954) and Gupta (1961) later 
found that it in fact varied with leaf size. 

The most variable venation character with regard to the present study is the 
degree to which the various orders of veins are represented on the epidermis. This 
feature appears as the epidermal cells are enlarging, but its exact time of develop- 
ment probably varies from group to group. In Buchenavia capitata and B. kleinii, 
for instance, it develops earlier on the lower epidermis, where it is quite well developed 
well before the cells are full-sized, than on the upper epidermis, where it has not 
fully developed when the leaves are fully grown : in the latter case both the smaller 
venules and undulations of the cells develop as the full-sized leaf matures. In fact 
it is the non-appearance of undulations on the venule cells which causes them to be 
distinguished so easily. 

No comments concerning the phenotypic variation of the epidermal veins seem 
to have been made in the literature, but considerable variation was in fact found in 
a number of species. Since the species in which this was noted were all of tropical 
origin the reasons could not be definitely explained, but| there is little doubt that 
they are partly environmental and positional in origin. The differences were 
concerned with the size of the smallest venules distinguishable on the epidermis, 
and thus with the sizes of the areolae and the degrees to which the venous system 
appeared to be open or closed. Usually the range of variation within a species 
gives extremes with different-sized areolae only, but in some cases a species shows 
open and closed venous systems of all degrees. 

It appears that the environmental conditions producing a leaf with fewer veins 
discernible on the epidermis in fact act primarily by producing a thicker leaf due to 
the larger cells and inter-cellular spaces. This in some way seems to permit fewer 
of the vein extensions to extend from the venules as far as the epidermis, so that 
fewer venules produce epidermal cell modifications. Greater shade and humidity 
are the commonest conditions which cause these effects. 

F. Development of cuticular membrane. 

The mature thickness of the cuticular membrane is not attained until a fairly late 
stage, when cell enlargement is completed. The membrane is, however, thick 
enough to be isolated in toto even before cell division has stopped, and the cuticular 
flanges are then well enough developed to show quite clearly the outlines of the cells. 
Nevertheless absolutely mature leaves are essential for comparative taxonomic 
purposes. 

Furthermore environmental conditions are known to affect the cuticular membrane 
thickness and cuticular-flange length. Skoss (1955) found that the " cuticle " was 
thicker in leaves exposed to the sun, and the same was discovered by a number of 
other workers who also correlated a thicker cuticular membrane with a drier soil, 
drier air, more exposure to wind, a higher altitude and other factors [fide Odell, 1932). 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 67 

Also the cuticular membrane is usually thicker on the upper stem leaves than on 
the lower stem or radical leaves (Yapp, 1912) but Stober (1917) reported that in 
some plants it was thickest on the lower stem and upper radical leaves. This 
cuticular thickness is accompanied by a thickening of the outer epidermal cellulose 
cell wall. 

Variation may also be caused by other, unknown, conditions, since Stevens (1932) 
noted a fluctuation in the cuticular thickness on cranberry fruits {Vaccinium 
macrocarpon Alton) from year to year which was not correlated with fruit size or 
climate, and which was exactly repeated in 33 different varieties of the plant. In 
Lumnitzera racemosa, a mangrove with a normally very thick cuticular membrane, 
a single specimen was encountered with very abnormally thin leaves due to the 
sparse development of the usually abundant water-storage tissue. This specimen 
had a very much thinner cuticular membrane than usual, but the epidermal anatomy 
was otherwise quite typical, and the habitat gave no clue to the reason for the 
pecuharity. 

Conspicuous cuticular striations are absent from most leaves, but when they are 
present they do not develop until a relatively late stage, and their usefulness is 
restricted to mature leaves. They are, however, of great taxonomic value in some 
groups, being highly characteristic of three of the four species of Macropteranthes, 
for example. No very young leaves of markedly striate species of the Combretaceae 
were available, but Martens (1934) has followed the development of the cuticular 
striations on certain petals. There is, however, a considerable variation in the 
degree of development of the cuticular striations in some species, and full account 
of this must be taken. In Combretum molle R. Br. ex G. Don, for example, the only 
known gathering from South West Africa is most distinctive in its extremely 
conspicuously striated cuticle, the striations being so strong as to obscure the cell 
outlines (Stace, 1961). In tropical Africa, however, where this species is very 
abundant, aU degrees of striation are found from the previous situation to the 
presence of only a few striations placed radially to each trichome-base, a situation 
present in many or even most species of the genus. Since the cell outlines in the 
latter cases are very clear, the two extremes have an entirely different aspect. The 
cause of this variation is, of course, unknown, but would seem likely to be environ- 
mental as the climate of the South West African habitat is relatively very dry. 

G. Summary. 

It is clear that a great deal of variation is to be encountered within one species in 
certain cuticular characters, although little or none may be found in others. Varia- 
tion may be attributed to minor genetic (genotypic) differences or to certain 
measurable influences, both genotypic and phenotypic, of three main types- 
developmental, positional and environmental. Obviously variation due to all of 
these factors must be fully considered before the characters are utilized taxonomically. 

In many or most plants the lower leaves are in a different environment from the 
upper, and their structure often reflects this. In Filipendula ulmaria, for example, 
the lower leaves are in a much more shaded and humid situation, and the type of 



68 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

variation from the upper leaves which they show is similar to that which is shown 
by upper leaves of other plants which are in a wholly damper environment, although 
this is not to say that the characteristics of the lower leaves are actually caused by 
their environment, as are those of upper leaves of plants in damp, shaded situations. 
The characters of these leaves are nevertheless to be regarded as relatively meso- 
morphic, and those of leaves in drier conditions as xeromorphic. The commonest 
mesomorphic cuticular characters are : a thinner cuticular membrane and shorter 
flanges ; more undulate epidermal cell walls ; fewer hairs ; less densely situated 
stomata ; larger epidermal cells and stomata ; and larger areolae. 

A third level of differentiation of mesomorphic and xeromorphic characters, 
beyond various parts of one plant and plants of one taxon, is between different 
species of one larger taxon. In this case the characters are genetically regulated, 
and they differ to some extent from those mentioned above, mainly in that they are 
concerned with structure to a greater extent than size and frequency, but they do 
show a number of similarities. The xeromorphic characters as exemplified by the 
Combretaceae are : a thicker cuticular membrane and longer cuticular flanges ; 
straighter cell walls ; more hairs (in a group in which hairs are of frequent occurrence 
the more xeromorphic species are usually the most pubescent, but in other groups 
xeromorphs are frequently extremely glabrous, e.g. mangroves) ; lesser development 
of veins and venules on the epidermis ; presence of stomata on both leaf surfaces 
(in those groups in which there is a tendency for stomata to appear on the upper 
epidermis the character is usually most developed in the most xeromorphic species, 
but in other groups the stomata of xeromorphs are frequently more limited in 
distribution on the lower epidermis alone, e.g. they may be confined to grooves or 
crypts) ; more sunken guard-cells ; and, within the anomocytic stomatal group, 
the tendency to develop distinctive epidermal cells adjacent to the guard-cells. 

It is the task of the taxonomist to distinguish between phenotypic and genotypic 
variation. In the former case the problem is to correlate it with specific environ- 
mental or other factors and to ascertain the degree of the variation, and in the latter 
to decide at what taxonomic levels it should be recognized, if at all. 

§8. General conclusions 

It cannot be too heavily emphasized that there are no guiding principles enabling 
assessment of the taxonomic usefulness of a given character in a taxon in which it 
has not been previously utihzed ; neither are there any means of predicting the 
characters most likely to prove of greatest diagnostic importance in an uninvestigated 
group. It is thus unwise and unjustifiable to draw on experience in one taxon 
and to apply it to another, even if the two plant groups are very closely related. 
In the Combretaceae, for instance, the flowers and trichomes provide the most 
useful systematic characters in the tribe Combreteae, the fruits being relatively 
uniform in structure ; whereas in the closely related Terminalieae the fruits and the 
epidermis are the only organs showing a large number of known taxonomic criteria. 
The diagnostic value of pollen-grains, chromosomes and wood anatomy in this 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 6g 

family has as yet been little studied. Furthermore, in general, characters tend to 
vary and assume a greater or lesser taxonomic importance quite independently of 
one another, and there is no single character which universally or even usually 
overrides all others in importance. Even in the case of the chromosomes, for which 
the latter has most frequently been claimed, there are very many taxa in which 
characters other than the number and morphology of these organelles are of greatest 
significance, indicating that the genetic information in the chromosomes is no more 
manifested in their over-all shape than in that of the flowers, fruits, leaves or 
other organs. 

There is, therefore, no reason to doubt the validity of epidermal features merely 
because in some groups they do vary greatly or do not agree with differentiation by 
other means. If it is possible for two taxa to differ solely on the shape of the stamens 
or leaves then it is possible for two others to be distinguished only on the form of 
their hairs or epidermal cells. It is very doubtful, however, whether a difference in a 
single feature of one organ (whatever that organ) will differentiate between taxa at 
a higher level than " minor genetic variants ", and almost certainly not between 
" species " as normally understood. In reaching taxonomic conclusions evidence 
from as many sources as possible should be incorporated into the catalogue of 
information which will lead to the proposed classification. No facet of the plant, 
morphological, physiological, biochemical or genetic, should be omitted. This 
becomes abundantly clear when the case of taxa whose systematic position varies 
with the character selectively utilized (e.g. Pteleopsis and Calycopteris in the Com- 
hretaceae) is considered. 

Undoubtedly minor exceptions to these generalizations exist, though by no means 
disproving the rule. There does appear, for instance, to be some degree of cor- 
relation between the diagnostic value of the epidermal cells and stomata and the 
degree of xeromorphy of the taxon. Most groups which have been found to possess 
taxonomically valuable epidermal characters are xeromorphic to some extent, e.g. 
gjminosperms, Lauraceae, Magnoliaceae, Ericaceae, Epacridaceae, various mangroves, 
and others. This is perhaps an expression of the idea that plants which are more 
highly specialized tend to be less variable, potentially as well as actually. It seems 
obvious that the greater the complexity of an organ the more features there are 
to be used by taxonomists, and thus the greater the diagnostic value of that organ. 
There is, in addition, the commonly held belief that characters of apparently little 
or no differential survival value (e.g. pollen-grain sculpturing, delimitation of 
cells of certain trichomes) are those with the greatest systematic and phylogenetic 
significance. 

Attempts have often been made to generalize on the taxonomic level at which 
various diagnostic characters differentiate. Fritsch (1903), for example, concluded 
that gland anatomy was more indicative above the generic level, and the presence 
or absence of papillate epidermal cells at the specific level. Edwards (1935), how- 
ever, pointed out that in different groups stomatal frequency can be a family, 
generic or specific criterion, or of no systematic importance at all. This is true of 



70 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

almost every character as the many exceptions to Fritsch's and others' claims 
testify. Thus, apart from noting that in a particular family or other group certain 
characters are more important than others at the various taxonomic levels little or 
no valid generalization can be made. A good example illustrating this point is 
the possession of a very densely or very sparsely pubescent leaf epidermis : in 
Ramatuella it serves to distinguish between R. argentea Kunth and R. virens Spruce 
ex Eichl. ; in Hibiscus vitifolius L. between subsp. vitifolius and subsp. vulgaris 
Brenan & Exell ; in Conocarpus erectus between var. sericeus and var. erectus ; and 
in many species of Epilohium between no more than various phenotypic modifications. 
In these cases, therefore, the level at which the character differentiates is presumably 
dependent upon the other characters with which it is correlated, and with the causes 
of the differences. Similarly, the presence of stomata on the upper epidermis is 
important at the family level within the mangrove habit, but amongst the non- 
mangrove members of the Comhretaceae it is of little significance, varying from 
species to species or often even infraspecifically. Stomatal size is frequently 
directly the result of the level of ploidy, and indicates a second reason for the fact 
that characters differentiate at varying taxonomic levels : namely that taxonomists 
have often failed to accord similar ranks to groups of taxa showing exactly the same 
differences. Thus Rorippa microphylla and R. nasturtium-aquaticum, Galium 
palustre L. subsp. palustre and subsp. tetraploideum Clapham, and Ranunculus 
ficaria var. ficaria and var. bulbifer Marsden- Jones are three pairs of taxa of which 
the first is diploid and the second tetraploid, although the resultant stomatal size 
differences apparently show differentiation at three distinct levels, if the taxonomic 
ranks assigned above are accepted^. 

The need to investigate exhaustively all aspects of the genotypic and phenotypic 
variation of a " new " character has been previously stressed. In cases where 
characters have been used before this preliminary survey has been undertaken the 
identifications attempted have often proved faulty, and have cut across decisions 
based on other characters. When the potentiality of the character is understood, 
however, and accurate identification is possible, the taxonomy of the group may on 
occasions be improved in the light of the discoveries. The present cuticular study 
has resulted in several such proposals in the classification of the Combretaceae. It is 
not to be expected that the new evidence will cause a drastic re-classification to 
become desirable, but rather that the new character might so shift the balance of 
the evidence from all sources that in some places taxonomic adjustments seem 
advisable. Entirely new classifications based on one type of evidence are usually 
shorter-lived than the original. 

Epidermal characters, then, are likely to provide not only a means of identification 
(especially of sterile material) but also as valuable taxonomic and phylogenetic clues 
as most other characters, and they exhibit much the same characteristics with 
respect to their pattern of variability. Edwards (1935) commented that the 

^ Rorippa microphylla was in fact first described as a variety of R. nasturtium-aquaticum, and Ranunculus 
ficaria var. bulbifer has been raised to subspecific rank as R. ficaria subsp. bulbifer (Marsden-Jones) 
Lawalrec. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 71 

cuticular differences between closely related species or infraspecific taxa are usually 
slight or absent. This is by no means always so. When all the differences between 
two closely allied taxa are analysed it is usually found that a few characters are 
conspicuously different, many slightly or scarcely so, and others identical. Thus, 
although it is most likely in such a case that the cuticular characters will also be 
only slightly or not different, the same is true of any other single feature. Occasion- 
ally, however, it is just that particular character being studied that shows the 
greatest difference between the two taxa. Several examples where this is true of 
the cuticular characters have been discovered within the Combretaceae. It seems 
that a knowledge of the relative usefulness and an understanding of the level of 
differentiation of cuticular characters in the angiosperms as a whole will only be 
objectively obtained when cuticular studies have been absorbed into the general 
practice of plant taxonomy. 

I do not claim that cuticular characters are of any outstandingly fundamental 
or all-important significance, as did some of the early exponents of the taxonomic 
use of pollen-grain or chromosome features. They are, however, to be regarded as 
characters of undoubted importance as further pieces of the jigsaw of complete 
systematic evidence, and at times they are in fact of greater value in identification 
and taxonomy than are any other characters of which we know. 



SPECIAL LITERATURE 

Ahmad, K. J. (1964). Cuticular studies in Solanaceae. Canad. Journ. Bot. 42 : 793-803. 

(1964a). On stomatal abnormalities in Solanaceae. Science &■ Culture 30 : 349-351. 

(1964b). Epidermal studies in Solanum. Lloydia 27 : 243-250. 

(1964c). Cuticular studies with special reference to abnormal stomatal cells in Cestrum. 

Journ. Ind. Bot. Soc. 43 : 165-177. 

Areschoug, F. W. C. (1902). Untersuchungen iiber den Blattbau der Mangrove-Pfianzen. 

Biblioth. Bot. 56. 
Arzt, T. (1934). Untersuchungen iiber das Vorkommen einer Kuticula in den Blattern 

dikotyler Pflanzen. Ber. Deutsch. Bot. Ges. 51 : 470-500. 
AsHBY, E. (1948). Studies in the morphogenesis of leaves. II. The area, cell size and cell 

number of leaves of Ipomoea in relation to their position on the shoot. New Phytol. 47 : 

177-195- 

& Wangermann, E. (1950). Studies in the morphogenesis of leaves. IV. Further 

observations on area, cell size and cell number of leaves of Ipomoea in relation to their 
position on the shoot. New Phytol. 49 : 23-35. 

& (1950a). Studies in the morphogenesis of leaves. V. A note on the origin of 

differences in cell size among leaves at different levels of insertion on the stem. New 
Phytol. 49 : 189-192. 

AvERY, G. S. (1933). Structure and development of the Tobacco leaf. Amer. Journ. Bot. 

20 : 565-592. 
Bachmann, O. (1886). Untersuchungen iiber die systematische Bedeutung der Schildhaare. 

Flora 69 : 387-400, 403-415, 428-448. 
Bailey, I. W. (1924). Notes on neotropical ant-plants. III. Cordia nodosa Lam. Bot. Gaz. 

77 : 32-49- 



72 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

Bailey, I. W., & Nast, C. G. (1944). The comparative morphology of the Winteraceae. 
V. FoUar epidermis and sclerenchyma. Journ. Arnold Arb. 25 : 342-348. 

& (1948). Morphology and relationships of Illicium, Schisandra and Kadsura. I. 

Stem and leaf. Journ. Arnold Arb. 29 : 77-89. 

Bandulska, H. (1923). A preliminary paper on the cuticular structure of certain dicotyle- 
donous and coniferous leaves from the Middle Eocene flora of Bournemouth. Journ. Linn. 
Soc. Lond., Bot. 46 : 241-269. 

(1924). On the cuticles of some recent and fossil Fagaceae. Journ. Linn. Soc. Lond., 

Bot. 46 : 427-441. 

(1926). On the cuticles of some fossil and recent Lauraceae. Journ. Linn. Soc. Lond., 

Bot. 47 : 383-425- 

{1928). A Cinnamon from the Bournemouth Eocene. Journ. Linn. Soc. Lond., Bot. 

48 : 139-147- 

(193 1). On the cuticles of some recent and fossil Myrtaceae. Journ. Linn. Soc. Lond., 

Bot. 48 : 657-671. 

Baranov, p. (1924). De la methode quantitative-anatomique de I'etude des plantes. I. 

Distribution des stomates. Bull. Univ. As. Centr. 7 : 30-35. 
Bary, A. DE (1871). Ueber die Wachsiiberziige der Epidermis. Bot. Zeit. 29 : 129-139, 

145-154, 161-176, 566-571, 573-585, 589-600, 605-619. 

(1884). Comparative anatomy of the vegetative organs of the phanerogams and ferns. Transl. 

F. O. Bower & D. H. Scott. Oxford. 

Bather, F. A. (1907). Nathorst's use of collodion imprints in the study of fossil plants. 

Geol. Mag., Dec. 5, 4 : 437-440. 
(1908), Nathorst's methods of studying cutinised portions of fossil plants. Geol. Mag., 

Dec. 5, 5 : 454-459- 
BoKORNY, T. (1882). Ueber die " durchsichtigen Punkte " in den Blattern. Flora 65 : 

339-350, 355-368, 371-381, 387-397. 411-417- 
BoRNEMANN, J. G. (1856). Ubcr organische Reste der Lettenkohlengruppe ThUringens. Leipzig. 
BoRRiLL, M. (1961). Epidermal characteristics in the diploid subspecies of Dactylis glomerata 

L. Journ. Linn. Soc. Lond., Bot. 56 : 453-458. 
Brenner, W. (1900). Untersuchungen an einigen Fettpflanzen. Flora 87 : 387-439. 
Brodie, p. B. (1842). Notice on the occurrence of plants in the Plastic Clay of the Hamp- 
shire coast. Proc. Geol. Soc. Lond. 3 : 592. 
Bureau, E. (1864). Monographic des BignoniacSes . Paris. 
BuRKiLL, I. H. (1939). Two notes on Dioscoreas in the Congo : (i) The acarodomatia of 

D. minutiflora Engl, and D. smilacifolia De Wild. Proc. Linn. Soc. Lond. 151 : 57-60. 
Carolin, R. C. (1954). Stomatal size, density and morphology in the genus Dianthus. 

Kew Bull. 9 : 251-258. 
Clarke, J. (i960). Preparation of leaf epidermis for topographic study. Stain Technol. 35 : 

35-39- 
CooKSON, I. C. (1953). On Macrozamia hopeites — an Early Tertiary cycad from Australia. 

Phytomorphology 3 : 306-312. 
Cowan, J. M. (1950). The Rhododendron leaf. Edinburgh & London. 
Dehnel, G. S. (i960). Response of stomata to wounding. Bot. Gaz. 122 : 124-130. 

(1961). Abnormal stomatal development in foliage leaves of Begonia aridicaulis. Amer. 

Journ. Bot. 48 : 129-133. 

DiLCHER, D. L. (1963). Cuticular analysis of Eocene leaves of Ocotea obtusifolia. Amer. 

Journ. Bot. 50 : 1-8. 
Edwards, W. N. (1935)- The systematic value of cuticular characters in recent and fossil 

angiosperms. Biol. Rev. &> Biol. Proc. Cambr. Phil. Soc. 10 : 442-459. 
Elliot, G. F. S. (1911). Botany of to-day. Ed. 2. London. 
Engler, a., & Prantl, K. (1887-1915). Die natiirlichen Pflanzenfamilien. Leipzig. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 73 

Esau, K. (1953). Plant anatomy. New York & London. 

ExELL, A. W. (1961). New species and subspecies of Combretum from southern tropical 
Africa. Mitteil. Bot. Staatssamml . Munch. 4 : 3-7. 

(1963). Plants of the 1957-58 West African Expedition : IV. A new species of Combretum 

from Northern Nigeria. Kew Bull, ij : 233-236. 

& Stage, C. A. (1963). A revision of the genera Buchenavia and Ramatuella. Bull. 

Brit. Mus. [Nat. Hist.), Bot. 3 : 1-46. 

Florin, R. (1920). Uber Cuticularstrukturen der Blatter bei einigen rezenten und fossilen 
Coniferen. Ark. for Bot. 16 (6). 

(1931). Untersuchengen zur Stammesgeschichte der Coniferales und Cordaitales. Erster 

Teil : Morphologic und Epidermisstruktur der Assimilationsorgane bei den rezenten 
Koniferen. K. Svensk. Vetenskapsakad. Handl., Ser. 3, 10 (i). 

(1933)- Studien iiber die Cycadales des Mesozoikums nebst Erorterungen iiber die 

Spaltoffnungsapparate der Bennettitales. K. Svensk. Vetenskapsakad. Handl., Ser. 3, 
12 (5). 

(1938-45). Die Koniferen des Oberkarbons und des Unteren Perms. Palaeontographica 

85 (B). 

Foster, A. S. (1949). Practical plant anatomy. Ed. 2. Toronto, New York & London. 
Fritsch, F. E. (1903). The use of anatomical characters for systematic purposes. New 

Phytol. 2 : 177-184. 
Goeppert, H. R., & Berendt, G. C. (1845). Der Bernstein und die in ihm befindlichen 

Pfianzenreste der Vorwelt. Berendt, Bernst. Org. Reste Vorwelt i (i). 
GoTHAN, W. (191 5). tJber die Methoden und neue Erfolge bei der Untersuchung kohlig 

erhaltener Pfianzenreste. Sitzungsber. Ges. Naturforsch. Freunde Berl. 1915 : 43-48. 
Grob, a. (i8g6). Beitrage zur Anatomie der Epidermis der Gramineenblatter. Biblioth. 

Bot. 36. 
Gupta, B. (1961). Correlation of tissues in leaves, i. Absolute vein-islet numbers and 

absolute veinlet termination numbers. Ann. of Bot., New Ser., 25 : 65-70. 
Hall, J. P., & Melville, C. (1951). Veinlet termination number. A new character for the 

differentiation of leaves. Journ. Pharm. cS' Pharmacol. 3 : 934-941. 

& (1954). Veinlet termination number — some further observations. Journ. 

Pharm. &- Pharmacol. 6 : 129-133. 

Harris, T. M. (1926). Note on a new inethod for the investigation of fossil plants. New 
Phytol. 25 : 58-60. 

(I942-*-). Notes on the Jurassic flora of Yorkshire, i^. Ann. &- Mag. Nat. Hist., 

Ser. 11,9: 568->-. 

(1944). Notes on the Jurassic flora of Yorkshire. 15. Ann. & Mag. Nat. Hist., Ser. 

II, II : 685-690. 

Heinen, W. (i960). tJber den enzymatischen Cutin-Abbau. I. Mitteilung : Nachweis 
eines " Cutinase "-Systems. Act. Bot. Neerl. 9 : 167-190. 

(1961). Uber den enzymatischen Cutin-Abbau. II. Mitteilung : Eigenschaften eines 

cutinolytischen Enzyms aus Penicillium spinulosum Thom. Act. Bot. Neerl. 10 : 1 71-189. 

Heintzelman, C. E., & Howard, R. A. (1948). The comparative morphology of the Icacina- 

ceae. V. The pubescence and the crystals. Amer. Journ. Bot. 35 : 42-52. 
HoFMANN, E. (1932). Epidermisreste und Blattabdriicke aus den Braunkohlenlagern des 

Geiseltales. Nov. Act. Leopold., Neue Folge, i : 59-74. 
Hohnel, F. von (1878). Einige Bemerkungen iiber die Cuticula. Oesterr. Bot. Zeitschr. 

28 : 81-87, 115-121. 
Holden, R. (1915). On the cuticles of some Indian conifers. Bot. Gaz. 60 : 215-227. 
Holtermann, C. (1893). Beitrage zur Anatomie der Combretaceen. Forhandl. Vidensk.-Selsk. 

Christiania 1893 (12). 
Hutchinson, J. (1926). The families of flowering plants. I. Dicotyledons. London. 

& Dalziel, J. M. (1931). Flora of west tropical Africa. 2 (i). London. 



74 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

J ALAN, S. (1962). The ontogeny of stomata in Schisandra grandiflora Hook. f. & Thorns. 

Phytomorphology 12 : 239-242. 
Jeffrey, E. C. (1917). The anatomy oj woody plants. Chicago. 
Johnson, T., & Gilmore, J. G. (1921). The occurrence of Dewalquea in the coal-bore at 

Washing Bay. Sci. Proc. R. Dublin Soc, New Ser., 16 : 323-333. 
Juniper, B. E. (1959). The surfaces of plants. Endeavour 18 : 20-25. 

(1959a). The effect of pre-emergent treatment of Peas with trichloracetic acid on the 

submicroscopic structure of the leaf surface. New Phytol. 58 : 1-4. 

(i960). Growth, development and effect of the environment on the ultra-structure of 

plant surfaces. Journ. Linn. Soc. Lond., Bot. 56 : 413-419. 

& Bradley, D. E. (1958). The carbon replica technique in study of the ultrastructure of 

leaf surfaces. Journ. Ultrastruct. Res. 2 : 16-27. 
Keay, R. W. J. (1953). Rhizophora in West Africa. Kew Bull. 8 : 121-127. 
Kostermans, a. J. G. H. (1957). Lauraceae. Reinwardtia 4 : 193-256. 
Kubart, B. (1927). Einige Bemerkungen zur Frage des Nachweises der Lauraceen-Gattung 

Umbellularia in den Tertiaren " Redwood "-Bestanden Nordamerikas. Biol. Gen. 3 : 

585-594. 
Lamarliere, L. G. de (1906). Sur les membranes cutinisees des plantes aquatiques. Rev. 

Gdn. Bot. 18 : 289-295. 
Lee, B., & Priestley, J. H. (1924). The plant cuticle. L Its structure, distribution, and 

function. Ann. of Bot. 38 : 525-545. 
Levin, F. A. (1929). The taxonomic value of vein islet areas. Quart. Journ. Pharm. 6^ 

Pharmacol. 2 : 17-43. 
LoFTFiELD, J. V. G. (1921). The behaviour of stomata. Carneg. Inst. Wash. Publ. 314. 
Long, F. L., & Clements, F. E. (1934). The method of collodion films for stomata. Amer. 

Journ. Bot. 21 : 7-17. 
LuNDSTROM, A. N. (1887). Pflanzenbiologische Studien. 11. Die Anpassungen der Pflanzen 

an Thiere. Nov. Act. R. Soc. Sci. Upsal., Ser. 3, 13 (10). 
McDougall, W. B. (1927). Plant ecology. London. 
Maheshwari, p., & Vasil, V. (1961). The stomata of Gnetum. Ann. of Bot., New Ser., 25 : 

313-319- 
Mani, M. S. (1964). Ecology of plant galls. Monogr. Biol. 12. 
Marlier-Spirlet, M. L. (1945). Sur quelques epidermes de feuilles de Cinnamomum L. 

Bull. Jard. Bot. Et. Brux. 17 : 266-305. 
Martens, P. (1934). Recherches sur la cuticle. II L Structure, origine et signification du 

relief cuticulaire. Protoplasma 20 : 483-515. 
Martin, D. J. (1955). Features on plant cuticle. An aid to the analysis of the natural diet 

of grazing animals, with esp)ecial reference to Scottish hill sheep. Trans. &- Proc. Bot. 

Soc. Edin. 36 : 278-288. 
Metcalfe, C. R. (1954). ^^ anatomist's view on angiosperm classification. Kew Bull. 

9 : 427-440- 
— — (1955)- Recent work on the systematic anatomy of the monocotyledons. Kew Bull. 
9 ■ 523-532. 

(i960). Anatomy of the monocotyledons. I.Gramineae. Oxford. 

(1963). Comparative anatomy as a modem botanical discipline. Advanc. Bot. Res. 

I : 101-147. 

& Chalk, L. (1950). Anatomy of the dicotyledons. Oxford. 

Meyer. F. J. (1932). Beitrage zur Anatomie der Alismataceen. Beik. Bot. Centralbl. 49 {i) : 
309-368. 

(1932a). Die Verwandtschaftsbeziehungen der Alismataceen zu den Ranales im Lichte 

der Anatomie. Engler, Bot. Jahrb. 65 : 53-59. 

Meyer, M. (1938). Die submikroskopische Struktur der kutinisierten Zellmembranen. 
Protoplasma 29 : 552-586. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 75 

MiNDEN, M. VON (1899). Beitrage zur anatomischen und physiologischen Kenntnis wasser- 

secernierender Organe. Biblioth. Bot. 46. 
MoRLEY, T. (1953). The genus Mouriri (Melastomaceae). A sectional revision based on 

anatomy and morphology. Univ. Calif. Publ. Bot. 26 : 223-299. 
Mueller, L. E., Carr, P. H., & Loomis, W. E. (1954). The submicroscopic structure of 

plant surfaces. Amer. Journ. Bot. 41 : 593-600. 
Nathorst, a. G. (1907-12). Palaobotanische Mitteilungen. i-ii. K. Svensk. Vetenskap- 

sakad. Handl., Ny Foljd, 42 (5)-48 (2). 

(1907a). t)ber die Anwendung von KoUodiumabdrucken bei der Untersuchung fossiler 

Pflanzen. Ark. for Bot. 7 (4). 

Netolitzky, F. (1932). Die Pfianzenhaare. Linsbauer, Handb. Pfianzenanat. 4 (29). 
North, C. (1956). A technique for measuring structural features of plant epidermis using 

cellulose acetate films. Nature 178 : 1186-1187. 
Odell, M. E. (1932). The determination of fossil angiosperms by the characteristics of their 

vegetative organs. Ann. of Bot. 46 : 941-963. 
Orgell, W. H. (1955)- The isolation of plant cuticle with pectic enzymes. PI. Physiol. 30 : 

78-80. 
Paliwal, G. S., & Bhandari, N. N. (1962). Stomatal development in some Magnoliaceae. 

Phytomorphology 12 ; 409-412. 
Pant, D. D. (1965). On the ontogeny of stomata and other homologous structures. PI. Sci. 

Ser. Allahabad i : 1-24. 

& Mehra, B. (1963). Development of caryophyllaceous stomata in Asteracantha longifolia 

Nees. Ann. of Bot., New Ser., 27 : 647-652. 

& Verm A, B. K. (1963). Development of stomata in leaves of Notonia grandifiora DC. 

Journ. Ind. Bot. Soc. 42 : 384-391. 

Paterson, B. R. (1961). Systematic studies of the anatomy of the genus Acrotriche R. Br. 

I. The leaf. Austral. Journ. Bot. 9 : 197-208. 
Penzig, O., & Chiabrera, C. (1903). Contributo alia conoscenza delle piante acarofile. 

Malpighia 17 : 429-487. 
Prantl, K. (1872). Die Ergebnisse der neueren Untersuchungen iiber die Spaltoffnungen. 

Flora 55 : 305-312, 321-328, 337-346, 369-382. 
Prat, H. (1932). L'epiderme des Graminees. Etude anatomique et syst^matique. Ann. 

Sci. Nat., Ser. 10, Bot. 14 : 117-325. 
Pray, T. R. (1963). Origin of vein endings in angiosperm leaves. Phytomorphology 13 : 

60-81. 
Priestley, J. H. (1943). The cuticle in angiosperms. Bot. Rev. 9 : 593-616. 
Prillieux, E. (1856). De la structure des polls des Oleacees et des Jasmin^es. Ann. Sci. 

Nat., Ser. 4, Bot. 5 : 5-14. 
Rao, H. S. (1939). Cuticular studies of Magnoliales. Proc. Ind. Acad. Sci., Sect. B, 9 : 

99-116. 
Rea, M. W. (1921). Stomata and hydathodes in Campanula rotundifolia L., and their relation 

to environment. New Phytol. 20: 56-72. 
Rehfous, L. (1914). Les stomates des Celastracees. Bull. Soc. Bot. Geneve, Ser. 2, 6 : 13-18. 

(1917)- Etude sur les stomates. These No. 605, Universite de Geneve. 

Reyenga, T. W., & Karstens, W. K. H. (1964). Changes in the epidermal pattern of the 
sepals of Hydrangea macrophylla (Thunb.) D.C., " Otaksa " in their successive growth 
phases. Formation of new stomata, disintegration of old stomata. Act. Bot. Neerl. 13 : 
340-351. 

Roelofsen, p. a. (1952). On the submicroscopic structure of cuticular cell walls. Act. 
Bot. Neerl. 1 : 99-114. 

(1959). The plant cell-wall. Linsbauer, Handb. Pfianzenanat., ed. 2, 3 (4). 

Salisbury, E. J. (1927). On the causes and ecological significance of stomatal frequency, 

with special reference to the woodland flora. Phil. Trans. R. Soc. Lond., Ser. B, 216 : 1-65. 



76 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 

Sax, H. J. (1938). The relation between stomata counts and chromosome number. Journ. 

Arnold Arb. 19 : 437-441. 
Sax, K., & Sax, H. J. (1937)- Stomata size and distribution in diploid and polyploid plants. 

Journ. Arnold Arb. 18 : 164-172. 
ScHENK, A. (1869-71). Beitrage zur Flora der Vorwelt. Ill, IV. Palaeontographica 19 : 

1 34, 203-276. 

ScHiEFERSTEiN, R. H., & LooMis, W. E. (1956). Wax deposits on leaf surfaces. PI. Physiol. 

31 : 240-247. 
& (1959)- Development of the cuticular layers in angiosperm leaves. Amer. Journ. 

Bot. 46 : 625-635. 
ScHLEiDEN, M. J. (1846). Ueber die fossilen Pfianzenreste des Jenaischen Muschelkalks. 

Schmid & Schleid., Geognost. Verhdltn. Saalth. Jena 2. 
ScHULZE, F. (1855). Bemerkungen iiber das Vorkommen wohlerhaltener Cellulose in Braun- 

kohle vind Steinkohle. Ber. Verhandl. K. Preuss. Akad. Wiss. Berl. 1855 : 676-678. 
Scott, F. M., Hamner, K. C, Baker, E., & Bowler, E. (1958). Electron miscroscope 

studies of the epidermis of Allium cepa. Amer. Journ. Bot. 45 : 449-461. 
Seithe, a. (i960). Die Haarformen der Gattung Rhododendron L. und die Moglichkeit ihrer 

taxonomischen Verwertung. Engler, Bot. Jahrb. 79 : 297-393. 
(1962). Die Haararten der Gattung Solanum L. und ihre taxonomische Verwertung. 

Engler, Bot. Jahrb. 81 : 261-335. 
Sifton, H. B. (1963). On the hairs and cuticle of Labrador Tea leaves. A developmental 

study. Canad. Journ. Bot. 41 : 199-207. 
Sinclair, C. B., & Dunn, D. B. (1961). Surface printing of plant leaves for phylogenetic 

studies. Stain Technol. 36 : 299-304. 
Skoss, J. D. (1955)- Structure and composition of plant cuticle in relation to environmental 

factors and permeability. Bot. Gaz. 117 : 55-72. 
Slade, B. F. (1957). Leaf development in relation to venation, as shown in Cercis siliquastrum 

L., Prunus serrulata Lindl. and Acer pseudoplatanus L. New Phytol. 56 : 281-300. 
Solereder, H. (1908). Systematic anatomy of the dicotyledons. Transl. L. A. Boodle & F. E. 

Fritsch ; revis. D. H. Scott. Oxford. 
& Meyer, F. J. (1928-33). Systematische Anatomie der Monokotyledonen. (Incomplete.) 

Berlin. 
Srivastava, p. N. (1956). Studies in the Glossopteris flora of India — 4. Glossopteris, 

Gangomopteris and Palaeovittaria from the Raniganj Coalfield. The Palaeobotanist 5 : 

1-45- 
Stace, C. A. (1961). Cuticular characters as an aid to the taxonomy of the South- West 

African species of Combretum. Mitteil. Bot. Staatssamml. MUnch. 4 : 9-19. 
(1963). Cuticular patterns as an aid to plant taxonomy. Ph. D. thesis (London). Ined. 

(1965). The significance of the leaf epidermis in the taxonomy of the Combretaceae. 

I. A general review of tribal, generic and specific characters. Journ. Linn. Soc. Lond., 
Bot. 59 : 229-252. 

Stebbins, G. L., & Jain, S. K. (i960). Developmental studies of cell differentiation in the 
epidermis of monocotyledons. I. Allium, Rhoeo, and Commelina. Developmental Biol. 

2 : 409-426. 

& Khush, G. S. (1961). Variation in the organization of the stomatal complex in the 

leaf epidermis of monocotyledons and its bearing on their phylogeny. Amer. Journ. Bot. 

48 : 51-59- 

& Shah, S. S. (i960). Developmental studies of cell differentiation in the epidermis of 

monocotyledons. II. Cytological features of stomatal development in the Gramineae. 
Developmental Biol. 2 : 477-500. 

Steenis, C. G. G. J. VAN (1953). Results of the Archbold Expeditions. Papuan Nothofagus. 
Journ. Arnold Arb. 34 : 301-374. 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 77 

Stevens, N. E. (1932). Thickness of cuticle in Cranberry fruits. Amer. Journl Bot. 19 : 

432-435- 
Stober, J. P. (191 7). A comparative study of winter and summer leaves of various herbs. 

Bot. Gaz. 63 : 89-109. 
Stockmans, F. (1932). Sur des epidermes de dicotyledonees {Dewalquea gelindenensis Saporta 

et Marion et Litsea elatinervis Saporta et Marion) des Marnes de Gelinden (Paleocene). 

Bull. Mus. R. Hist. Nat. Belg. 8 (9). 
Straus, A. (1930). Dikotyle Pflanzenreste aus dem Oberpliozan von Willershausen (Kreis 

Osterode, Harz) i. Jahrb. Preuss. Geol. Landes. Berl. 51 (i) : 302-336. 
SuRANGE, K. R., & Srivastava, P. N. (1956). Studies in the Glossopteris flora of India^5. 

Generic status of Glossopteris, Gangomopteris and Palaeovittaria. The Palaeobotanist 5 : 

46-49. 
Tateoka, T., Indue, S., & Kawano, S. (i959). Notes on some grasses. IX. Systematic 

significance of bicellular microhairs of leaf epidermis. Bot. Gaz. 121 : 80-91. 
Thomas, H. H. (1912). On some methods in palaeobotany. New Phytol. 11 : 109-114. 
(1930). Further observations on the cuticle structure of Mesozoic cycadean fronds. 

Journ. Linn. Soc. Lond., Bot. 48 : 389-415. 
& Bancroft, N. (1913). On the cuticles of some recent and fossil cycadean fronds. 

Trans. Linn. Soc. Land., Ser. 2, Bot. 8 : 155-204. 
Thompson, W. P. (191 2). Structure of the stomata of certain Cretaceous conifers. Bot. Gaz. 

54 : 63-67- 
TiMMERMAN, H. A. (1927). Stomatal numbers : their value for distinguishing species. Pharni. 

Journ. &- Pharmacist 118 : 241-243. 
Tomlinson, p. B. (1959). Anatomical approach to the classification of the Musaceae. Journ. 

Linn. Soc. Land., Bot. 55 : 779-809. 

(1961). Anatomy of the monocotyledons. //. Palmae. Oxford. 

TuRRELL, F. M. (1942). A quantitative morphological analysis of large and small leaves of 

Alfalfa, with special reference to internal surface. Amer. Journ. Bot. 29 : 400-415. 
Unger, F. (1853). Iconographia plantarum fossilium. Abbildungen und Beschreibungen 

fossiler Pflanzen. Denkschr. Kaiserl. Akad. Wiss. Wien, Math.-nat. CI. 4 (i) : 73-118. 
Uphof, J. C. T. (1942). Ecological relations of plants with ants and termites. Bot. Rev. 8 : 

563-598. 
Vesque, J. (1889). Ue I'emploi des caracteres anatomiques dans la classification des vegetaux. 

Bull. Soc. Bot. France 36 : xli-lxxvii. 
Watson, L. (1962). The taxonomic significance of stomatal distribution and morphology in 

Epacridaceae. New Phytol. 61 : 36-40. 
(1965). The taxonomic significance of certain anatomical variations among Ericaceae. 

Journ. Linn. Soc. Land., Bot. 59 : 111-125. 
Watson, R. W. (1942). The effect of cuticular hardening on the form of epidermal cells. 

New Phytol. 41 : 223-229. 
Weiss, A. (1865). Untersuchungen iiber die Zahlen- und Grossenverhaltnisse der Spaltoft- 

nungen. Jahrb. Wiss. Bot. 4 : 125-196. 
Wessel, P., & Weber, O. (1856). Neuer Beitrag zur Tertiarflora der Niederrheinischen 

Braunkohlenformation. Palaeontographica 4 : 111-178. 
Wildeman, E. de (1904). Sur I'acarophytisme chez les monocotyledones. Compt. Rend. 

Acad. Sci. Par. 139 : 551-553. 
(1923). Notes sur le Strephonema gilleti De Wild, plante oleagineuse du Congo. Bull. 

Jard. Bot. Et. Brux. 8 : 1 19-124. 
Wills, L. (1914). Plant cuticles from the Coal-measures of Britain. Geol. Mag., Dec. 6, 

I : 385-390. 



78 



CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 



Wylie, K. B. (1943). The role of the epidermis in foHar organization and its relations to the 

minor venation. Amer. Journ. Bot. 30 : 273-280. 
Yapp, R. H. (191 2). Spiraea ulmaria, L., and its bearing on the problem of xeromorpliy in 

marsh plants. Ann. of Bot. 26 : 815-870. 
Zp:iller, R. (1882). Observations sur quelques cuticules fossiles. Ann. Sci. Nat., Ser. 6, 

Bot. 13 : 217-238. 



\ 



Bull. B.M. (N.H.) Bot. 4. i 





A, Rhizophora mucronata Lam. : lower epidermis of leaf, showing cork-wart (X420). 
B, Buchenavia kleinii Exell : upper epidermis of leaf, showing wound caused by fungal 
attack ( X420). C, Terminalia catappa L. : leaf margin, showing characteristic hole 
( X420). D, Lumnitzeva racemosa Willd. : lower epidermis of leaf, showing water-stoma 
(X420). 



\ 



Bull. B.M. (N.fl.) Hot. 4, i 



PLATE 2 








,">;*'4ii£M 




D 



A, Buchenavia grandis Ducke : lower epidermis of leaf, showing marsupiform domatium 
(x8o). B, B. capitata (Vahl) Eichl. : lower epidermis of leaf, showing marsupiform 
domatium ( x 6o). C, Terminalia catappa L. : lower epidermis of leaf, showing lebetiform 
domatium ( x8o). D, Conocarpns erectus var. sericeiis Forsstr. ex DC. : lower epidermis 
of leaf, showing lebetiform domatium ( x6o). 



\ 



Bull. B.M. (N.H.) Bot. 4, i 



PLATE 3 




A, Rhizophora mangle L. : lower epidermis of leaf, showing strongly cuticularized stomata 
and straight epidermal cell walls ( X420). B, Buchenavia fanshawei Exell & Maguire : 
lower epidermis of leaf, showing thinly cuticularized stomata and undulate epidermal 
cell walls ( X 420). C.Combretum nigrescens'Kmg : leaf margin ( x 170). D, Lagunailaria 
racemosa (L.) Gaertn. f. : leaf margin ( X170). 



\ 



Bull. B.M. (X.II.) Hot. 4, I 



PLATE 4 



.!■ 






> 




-'^ 




,/:#^i 




:3 




^^^^^Kl^^tlSfl^^^^l 


^^^^^^^H 


1 . ^ •v^rt.'*^ 




^^i 


^^^^ 






'^1 


i 


# ^ 


::» Mi Mm 






%n-. . 


( 


L; 




t 




s-.^^ 





i A 







/ ■"♦* 





A, Ramatuella virens Spruce ex Eichl. : lower epidermis of leaf, showing anomocytic stomata 
(X420). B, Laguncularia racemosa (L.) Gaertn. f. : lower epidermis of leaf, showing 
cyclocytic stomata ( X420). C, Strephonema sericeum Hook. f. : lower epidermis of leaf, 
showing paracytic stomata ( X420). D, Guiera senegalensis J. F. Gmel. : upper epidermis 
of leaf, showing numerous simple hair-bases ( X420). 



\ 



Bull. B.M. {N.H.) Bot. 4, i 



PLATE ^ 









■^/^ 



, '^"-,.6'^ '1^-,. 








V , 








# 



i^ 



"H 
.' A 






A, Thiloa glaucocarpa (Mart.) Eichl. : lower epidermis of leaf, showing scale with strongly 
cuticularized scale-base ( X420). B, Combretimi obovatum F. HofJm. : lower epidermis of 
leaf, showing stalked glands on the raised venule reticulum ( X170). C, Ceriops tagal 
(Perrott.) C. B. Robinson : lower epidermis of leaf, showing stoma and cuticular flanges 
in profile on cuticular membrane folded inside outermost ( X420). D, Ouisqiialis hensii 
(Engler & Diels) Exell : upper epidermis of leaf, showing cuticular striations and three 
types of trichome-base ( X420). 



D 



\ 



i 




■',^?'*?*vS 






■f'^^'Sfc" 



'. ", Vv' . ■ • • ■■■■' ■ , ::^-^.;.^;^;^0- -^^rV".- --n^r^^ 



GREAT BRITAIN 
DE LA RUE & 
VNY LIMITED LONDON 



'7 



THE GENUS ELAPHOGLOSSUM 

IN THE INDIAN PENINSULA 

AND CEYLON 



W. A. SLEDGE 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

BOTANY Vol. 4 No. 2 

LONDON: 1967 












iiiiiftiii^ii'iiailmii^^ 



1 MAY 1967 



THE GENUS ELAPHOGLOSSUM IN THE \^^^ ^^^^ 
INDIAN PENINSULA AND CEYLON 



BY 

W. A. SLEDGE 



(University of Leeds) 



Xa 



Pp. 79-96 



BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

BOTANY Vol. 4 No. 2 

LONDON : 1967 



THE BULLETIN OF THE BRITISH MUSEUM 

(NATURAL HISTORY), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 2 of the Botany 
series. 



World List abbreviation 
Btdl. Br. Mus. nat. Hist. (Bot. 



© Trustees of the British Museum (Natural History) 1967 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 2 May, 1967 Price Seven Shillings 



THE GENUS ELAPHOGLOSSUM IN THE 
INDIAN PENINSULA AND CEYLON 

By W. A. SLEDGE 

In a comment on the systematic list of ferns in a paper by Manton and myself on the 
cytology and taxonomy of the pteridophytes of Ceylon, I expressed the view 
(Phil. Trans. Roy. Soc, ser. B. 238 : 158 (1954)), that the names currently employed 
for the four species of Elaphoglossum were in need of revision. It subsequently 
became apparent that both the taxonomy and the nomenclature of the Indian 
species also required re-examination. 

The names employed in Beddome's works are mostly those originally given to 
African or tropical American plants with which Indian ones were at that time 
considered identical. The same names are now used in a more restricted sense and 
the discrimination of several species with more limited distributions within the areas 
of the aggregate species has left in doubt the correct names to be assigned to Indian 
and Ceylon species. So far as I am aware the only paper on Indian species of 
Elaphoglossum which has appeared since the time of Beddome is one by Biswas 
(Bull. Misc. Inf. Kew 1939 : 237-241 (1939)) in which two new species are described. 
Both of these supposed new species however are unnecessary renamings of species 
already described by Blume and Fee. 

In his Ferns of Southern India, Beddome described under the name Elaphoglossum 
conforme a species which appears to be not uncommon in the Nilgiri Hills. Biswas, 
who described it as new, was correct in rejecting this identification but appears 
not to have questioned the identification of another species growing in the 
Nilgiri and adjoining hills which Beddome and later authors called E. stigmatoUpis 
(Fee) Moore ; indeed he makes no reference anywhere in his paper to this species. 
Beddome's E. conforme was in fact based on specimens of the same species as that 
first described by Fee as Acrostichum stigmatolepis , and it is therefore Beddome's 
E. stigmatolepis which requires a new name. In his Handbook to the Ferns of British 
India, Ceylon and the Malay Peninsula (1883), Beddome amplified the distribution 
credited to E. conforme to include the Himalayas, Ceylon and Malaya. The plant 
from the Himalayas is E. marginatum, which had long been regarded as a synonym of 
E. conforme but which Biswas correctly reinstated as a distinct species. Neither 
E. stigmatolepis nor E. marginatum occur either in Ceylon or Malaya and these regions 
were included through errors of identification and do not cover a third species un- 
accounted for amongst those described by Beddome. 

The Indian specimens which Beddome originally referred to E. laurifolium (changed 
in his Handbook to E. latifolium) are identical with Blume's Javan Acrostichum 
angulatum, which name he cited as a synonym of E. laurifolium in his Ferns of 
Southern India. This is the other species which Biswas needlessly described as a 
new species. E. angulatum is an easily recognized species on account of its long- 
creeping rhizome and very distinctive rhizome scales, yet most Ceylon gatherings of 

BOX. 4, 2. 6§ 



82 ELAPHOGLOSSU M IN THE INDIAN PENINSULA AND CEYLON 

this species are identified in herbaria not as E. laurifolium but as E. conforme. 
The true Acrostichum laurifolium of Thouars is confined to Tristan da Cunha and 
Gough Island, and another species from Ceylon and southern India with short- 
creeping rhizome, closely placed fronds and different rhizome scales, which is usually 
labelled E. laurifolium in herbaria, is E. commutatum, a species which was founded by 
Mettenius partly on a Ceylon gathering made by Thwaites but which is not illustrated 
in Beddome's works. 

As regards the ferns that Beddome called E. squamosum, this name was currently 
used to cover superficially similar plants ranging from tropical America eastwards 
to southern India and Ceylon. African, Madagascan and Mascarene representatives 
of the aggregate species have since been split off from Jamaican plants, whence the 
type was described. The Nilgiri and Ceylon plants which were united under the 
same name by Beddome are certainly distinct from one another and as neither 
can be matched by African or Mascarene gatherings, I have described these as two 
new species, using names already given to herbarium specimens at Kew by Krajina 
but never published by him. 

Of the names used by Beddome, only E. spatulatum stands unchanged and even 
this very distinct species has recently been incorrectly excluded as a Ceylon species 
by Madame Tardieu-Blot (Not. Syst. 15 : 432 (1959)), and the equally distinct Mada- 
gascan E. schizolepis has erroneously been substituted for it. 

The confusion which has surrounded the application of so many of these names is 
discussed in more detail later in this paper under the individual species ; but it may 
be useful to list here the illustrations and names given in Beddome's Ferns of Southern 
India together with their correct identifications : 

F.S.I, t. 196 as E. viscosum = E. stelligerum (Wall, ex Baker) 

Alston & Bonner 
F.S.I, t. 197 as E. squamosum — E. nilgiricum Krajina ex Sledge 
F.S.I, t. 198 as E. conforme = E. stigmatolepis (Fee) T. Moore 
F.S.I, t. 199 as £. stigmatolepis = E. beddomei Sledge 
F.S.I, t. 200 as E. laurifolium =■ E. angulatum (Blume) T. Moore 
F.S.I, t. 213 as E. spathulatum = E. spatulatum (Bory) T. Moore 

E. marginatum (Wall, ex Fee) 

T. Moore 
E. commutatum (Mett.) Alderw. 

van Rosenb. 
E. ceylanicum Krajina ex Sledge 

At the time of writing, a proposal (Anderson in Regn. Veg. 40 : 18 (1965)) to con- 
serve the generic name Elaphoglossum Schott ex J. Smith (Hooker's Journ. Bot. 
4 : 148 (Aug. 1841)) a.gaimst A coniopteris C. Presl (Tent. Pterid. : 236 (1836)) is under 
consideration. It has been approved by the Special Committee for Pteridophyta 
(Taxon 15 : 333 (1966)) but awaits ratification. 

My grateful thanks are tendered to the Directors and Curators of the herbaria at 
the British Museum (Natural History) (BM), the Botany School, University of 
Cambridge (CGE), the Royal Botanic Garden, Edinburgh (E), the Royal Botanic 



No illustration in F.S.I. 



ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 83 

Gardens, Kew (K) and the Division of Systematic Botany, Peradeniya, Ceylon 
(PDA) for access to, or for the loan of, specimens in their charge. 

Key to the Species 

Fronds coriaceous or at least stiff in texture, margins with a cartilaginous border, 
surfaces glabrous or with minute scales. 
Rhizome long-creeping, fronds spaced. 

Rhizome scales papery, broad, pale brown ; lamina 2'5-4-5 cm. broad 

1. E. angulatum 
Rhizome scales not papery, dark brown or black ; lamina 1-5-2 -5 cm. broad 

2. E. stigmatolepis 
Rhizome short-creeping, fronds crowded. 

Rhizome scales entire, linear (5-10 times as long as broad), stipes much 
shorter than lamina . . . . . . 3. £. heddomei 

Rhizome scales more or less fimbriate, narrow (4-5 times as long as broad), 
stipes up to half as long as lamina. 
Rhizome scales dark brown, 5 mm. long, acuminate 4. E. marginatum 

Rhizome scales brown, 10 mm. long, attenuate with crisped tips 

5. £■. commutatum 
Fronds thickly herbaceous or soft in texture, margins without a cartilaginous border, 
surfaces conspicuously scaly or hairy. 
Scales of stipe and lamina stellate with long rays . . 6. £. stelligerum 

Scales not stellate. 

Scales ovate with long marginal teeth. 

Fronds densely paleate, the lower surface wholly obscured by imbricating 
scales ; scales of stipe broad, mostly concolorous y. E. nilgiricum 

Fronds paleate but neither surface obscured by scales ; scales of stipe 
narrow, mostly with their borders and teeth dark-coloured 

8. £'. ceylanicum 
Scales fiHform, hair-like . . . . . . 9. £. spatulatum 

I. Elaphoglossum angulatum (Blume) T. Moore, Index Fil. : 5 (1857). 

rAcrostichum angulatum Blume, Enum. PL Jav. : loi (1828) ; Fl. Jav., Fil. : 25, t. 6 (1829). — 
F6e, M6m. Fam. Foug. 2 : 32 (1845). — Kunze in Linnaea 24 : 248 (1851). — Christ in 

Neue Denkschr. Schweiz. Ges. Naturw. 36 (i, 2) : 46, 49 (1899). 
Olfersia angulata (Blume) C. Presl, Tent. Pterid. : 234 (1836). 
Acrostichum conforme sensu Hook., Sp. Fil. 5 : 198 (1864) pro parte ; non Swartz. — Hook. & 

Baker, Synops. Fil. : 401 (1868) pro parte. 
Acrostichum marginatum sensu Thw., Enum. PI. Zeyl. : 380 (1864) ; non Wall, ex F6e. 
Elaphoglossum laurifolium sensu Bedd., Ferns S. Ind. : 67, t. 200 (1864) ; non T. Moore. 
Elaphoglossum latifolium sensu Bedd., Handb. Ferns Brit. Ind. : 416 (1883) ; non J. Smith. 
Elaphoglossum ogatai C. Chr., in Dansk. Bot. Ark. g (3) : 67 (1937). 
Elaphoglossum krajinae Biswas in Bull. Misc. Inf. Kew 1939 : 240, t. i fig. 2 (1939) ■ 

Rhizome long-creeping with stipes 1-3 cm, apart, clothed with pale, rufous-brown, 
papery, thin-walled, broadly ovate, acute scales, 5x2 mm., with subentire margins. 
Fronds stalked 15-35 (45) cm., stipe of sterile fronds up to 15 cm. or more, often as 



84 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

long as the lamina but sometimes only half or less their length, scaly, the scales like 
those of the rhizome but with irregularly fringed margins ; lamina elliptic, 2-5-5 cm. 
broad, apex acute, base narrowed and shortly decurrent on the stipe, margins with a 
cartilaginous, revolute border, lower surface with scattered, minute, brown, more or 
less stellately-laciniated scales, midrib with a few larger irregularly laciniate scales ; 
texture coriaceous. Fertile fronds on stalks about as long as or longer than the 
sterile ones, 2-3 cm. broad. Spores 27-30 X 18-21 /^. 

South India: Anamallays, 1,800 m., on trees, " F.S.Ind. f. 200 Elaphoglossum 
laurifolium" Beddome (BM). Kudiabad, Sispara Rd., 2,400 m., Nov. 1883, Gamble 
13473 (K). Bear Shola, Kodaikanal, Pulney Hills, 19 May 1898, Bourne 4gy2 (K). 
Kodaikanal Waterfall, Pulney Hills, 30 May 1898, Bourne 4gy3 (K). 

Ceylon : Thwaites C.P. 1311 (BM ; CGE ; K ; PDA). Gardner 1165 (BM ; 
CGE, Nuwara Eliya, on trees, 1,500-2,100 m., Sept. 1844 ; E ; K). Gardner 1310 
(E ; K). Nuwara Eliya, Thomson (K). Same locality. Freeman 36gA, 370B, 37 iC 
(BM). Same locality, Day in Herh. Henderson {E.). Same locality, creeping over 
tree trunks, 9 May 1906, Matthew (K). Pedrotalagala, 2,100 m., on mossy trees, 
26 Dec. 1950, Holttum (SING). Adams Peak, on trees, 2,100 m., 14 Feb. 1908, 
Matthew (K). Same locality, 1,950 m., 14 Dec. 1950, Sledge 615 (BM). Namunukula, 
Uva Province, 1,950 m., 24 Feb. 1954, Sledge 1208 (BM). 1855, R. W. Rawson lojg 
(BM). Mrs. Carr ex Herb. T. Moore (K). Wall m. Herb. Henderson dindm Herb. 
Neill Fraser (E). 

South India (Nilgiri, Pulney and Anamallay Hills), Ceylon, Sumatra, Java, 
Borneo, New Guinea, Philippine Islands, Tonkin, Formosa. 

The long-creeping rhizome and broad, membranous, pale brown or rufous-brown, 
concolorous scales readily distinguish E. angulatum from all other Asiatic species. 
It differs also from all other Indian or Ceylonese species in having the veins united 
by a connecting strand at their extremities — a character of the Aconiopteris {E. 
nervosum (Bory) Christ) group. This structural feature is however invisible in living 
and herbarium specimens and can only be seen after treatment with a clearing agent. 
Its spores are smaller than those of the other marginate-leaved species occuring in 
India and Ceylon. E. stigmatolepis , which it most resembles in growth habit, may be 
distinguished by its dark rhizome scales and the narrower fronds appearing punctate 
beneath on account of the more peltate scales and the black points of attachment 
left where these have been shed. 

Ceylon gatherings of this fern are usually named E. conforme and Indian ones 
either E. conforme or E. laurifolium (or E. latifolium). They certainly resemble 
African E. conforme but they agree far better and are indeed indentical with authentic 
specimens of Blume's A. angulatum at Kew and with other gatherings from Java. 
Kunze (loc. cit.) long ago reached the same conclusion when he referred a Nilgiri 
gathering {Schmid 65) to this species. A. angulatum was later merged in A. conforme 
by Hooker in the Species Filicum, and subsequent authors followed Hooker in their 
treatment of Ceylon and Indian plants. 

Beddome's illustrations of what he names E. conforme and E. laurifolium in his 
Ferns of Southern India and the descriptions in his Handbook have been the greatest 



ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 85 

source of confusion surrounding the taxonomy of Indian Elaphoglossums. In the 
former work his illustration (t. 198) and description, purporting to be ^. conforme 
Swartz, were based on plants on a sheet in the Kew collection annotated by him 
" type of tab. 198 F.S.I. ". The specimens on this sheet represent unmistakable 
though small and not very characteristic examples of E. stigmatolepis (Fee) Moore. 
(Beddome's illustration (t. 199) of what he calls E. stigmatolepis Fee represents another 
species.) The two different scales depicted represent examples taken from the 
rhizome and, in the case of the shorter and broader one with markedly jagged margins, 
from the stipe or midrib of the frond. The enlargement of a portion of the underside 
of the frond shows stellate scales quite different from the minute peltate scales with 
shortly fimbriate margins which characterize E. stigmatolepis, and this drawing could 
not possibly have been made from any of the fronds on the sheet referred to. There 
is moreover no reference in the accompanying description or in the later description 
in the Handbook to the presence of stellate scales on the fronds. The earlier work 
similarly contains no description of the form or colour of the rhizome scales, but in 
the Handbook these are referred to as " blackish " in colour, a description which is 
applicable to the specimens on the sheet and appropriate for E. stigmatolepis (and 
even, though somewhat less accurately, for E. marginatum, which, from the amplified 
distribution assigned in the Handbook to " conforme ", he now included in that species) 
but which is completely false for E. angulatum or for the true African E. conforme. 

It is clear from the evidence provided by Beddome's own specimens at Kew and 
the British Museum that he was unfamiliar with E. angulatum for, whereas E. 
stigmatolepis (i.e. Beddome's E. conforme) is stated to be " very common " and is 
represented by several gatherings of his, he appears to have collected E. angulatum 
once only. This single gathering is in the British Museum collection, the sheet being 
labelled by Beddome " Anamallays. 6,000 ft., on trees. Elaphoglossum laurifoli^im 
(Thouars) F.S.Ind. f. 200 ". The description of E. laurifolium in the Ferns of 
South. India is followed by the citation of " E. angulatum Bl. En. 201 [sic] : Id. Fl. 
Jav. 25, t. 6 " in synonomy and the reference to the rhizome as " wide-creeping with 
bright chestnut or golden scales " and the other distinctions attributed in the 
Handbook to his E. conforme and E. latifolium (the name there substituted for E. 
laurifolium of the Ferns of Southern India) become explicable when it is realized that 
his E. conforme is based on E. stigmatolepis and his E. laurifolium (or E. latifolium) 
is based on E. angulatum. 

Christensen states (in Gard. Bull. Str. Settl. 7 : 291 (1934)) a propos of Blume's 
figures of his A. conforme (i.e. E. commutatum) and A. angulatum, " I am by no 
means sure that the two ' species ' illustrated by Blume are really different ". And 
so it might seem from the plates themselves, for whilst the plate of the latter is an 
admirable representation of E. angulatum, that of the former is less characteristic 
of E. commutatum particularly in the spacing and small size of the fronds. Blume's 
drawings and descriptions of the rhizome scales of his two species however indicate 
such considerable differences and they agree so very closely with the scales of E. 
commutatum and E. angulatum that I have no hesitation in accepting the descriptions 
and the plates as illustrations of these two species. Fee's plate of A. laurifolium 
does not represent the same species. It is a Mascarene plant which has since been 



86 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

described by Mme. Tardieu-Blot (Not. Syst. 15 : 433 (1959)) as E. alstonii. 

The type of E. ogatai is at the British Museum. Of this Christensen (DanskBot. 
Ark, 9 (3) : 67 (1937)) says " it is extremely hke E. angulatum (Bl.) Moore . . . differs 
however in its remarkably thin texture and the distinctly but shortly decurrent 
lamina ". Tagawa, who is familiar with the living plant, states (in Mem. Coll. Sci. 
Univ. Kyoto, ser. B, 20 : 28 (1951)) that its fronds are thin-coriaceous and that no 
distinction can be drawn in the other respects. One of Bourne's Indian plants 
(Bear Shola, Pulney Hills) also has quite thin fronds, and the more ovate rather 
than elliptic fronds which look distinctive in the type specimen of E. ogatai can be 
matched by a specimen of Matthew's from Adam's Peak, Ceylon. I therefore accept 
Tagawa' s reduction of E. ogatai to a synonym of E. angulatum. 

Biswas's E. krajinae is quite clearly no more than a redescription of E. angulatum, 
which he evidently never took into consideration when separating Indian and Ceylon 
plants from African E. con forme. 

2. Elaphoglossutn stigmatolepis (Fee) T. Moore, Index Fil. : xvi, 15 (1857), 
368 (1862). — Christ in Neue Denkschr. Schweiz. Ges. Naturw. 36 (i, 2) : 52 
(1899). 

AcrosHchum stigmatolepis F6e, Mem. Fam. Foug. 2 : 62, t. 24 f. 2 (1845). — Hook., Sp. Fil. 

5 : 216 (1864). 
Elaphoglossum conforme sensu Bedd., Ferns S. Ind. : 67, t. 198 (1864) ; Handb. Ferns 

Brit. Ind. : 416, fig. 247 (1883) pro parte ; non J. Smith. 
Elaphoglossum ballardianum Biswas in Bull. Misc. Inf. Kew 1939 : 239, t. i fig. i (1939). 

Rhizome long-creeping, stipes 5-10 mm. apart, clothed with ovate-lanceolate scales, 
3-5 X 1-2 mm., brown at the base and blackish above, with subentire or sparsely 
fringed margins and acute tips. Fronds stalked, 15-35 cm. long, stipes of sterile 
fronds variable in length, sometimes as long as the lamina but commonly much less, 
scaly, at least when young, with rather broad scales with jagged margins ; lamina 
narrowly elliptic, i-5-2'5 cm. broad, apex acuminate, base narrowed and decurrent 
on the stipe, margins with a narrow, revolute, cartilaginous border, midrib sparsely 
scaly beneath with short broad scales like those on the stipe, lower surface clothed with 
minute, dark, peltate scales with fimbriate margins, upper surface with scattered scales 
of the same form when young, becoming smooth with age ; texture stiffly mem- 
branaceous to subcoriaceous. Fertile fronds on stipes normally exceeding those of 
the sterile fronds, 1-2 cm. broad. Spores (39) 42-45 (48) X 27-30 ju. 

South India : " Ind. Orient Neilgherries, 159 " (CGE, Pisotype). Nilgiris, 
2,100 m., " type of tab 198 F.S.I.", Beddome (K). Same locality, i860, Beddome 5 
(K). Same locality, 2,100 m., Oct. 1889, Gamble 21420 (K). Same locality, 2,100 m. 
Nov. 1870, Henderson (E) . Same locality, Wight 50 (K) . Same locality, on trees and 
rocks, 1842, Gough 3234 j^S (K). Ootacamund, 2,100 m., Oct. 1885, Gamble i6g63 
(BM). Same locality and date. Gamble in Herb. Blanford (E). Same locality, 2,100 
m., Aug. 1885, Gamble 16661 (K). Same locality, Babcock (K). Same locality, 
Cockburn 85 (BM). Dodabetta, Nilgiri District, 2,550 m., June 1883, Gamble 
12038 (E). Kudiabad, Sispara Road, 2,400 m., Nov. 1883, Gamble 13474 (K). 



ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 87 

Moir Point, Pulney Hills, 2,340 m., stream, July 1937 and May 1938, Aroticasamy 10 

(BM). 

South India (Nilgiri and Pulney Hills). 

Elaphoglossum stigmatolepis differs from the other marginate-leaved species in its 
narrower, less coriaceous fronds, which are more tapering above and studded beneath 
with minute disc-like scales with shortly fimbriate margins. The points of attach- 
ment to the lamina appear as black dots when the scales have been shed. These 
are referred to by Fee as "petits points qui paraissent etre de nature glanduleuse: 
ne serait-ce pas la base persistante des squames? ' ' . Young fronds show more sparsely 
distributed scales on the upper surface and in older fronds, where they have been 
shed, similar minute black scars are recognizable. Fee does not allude to these in his 
description but they are correctly shown in his illustration. 

The rhizome scales are more pointed than those of E. angulatum, pale brown at 
the base but almost black, and often somewhat glossy, in the upper parts, and hence 
through imbrication they appear darker than the uniformly pale, rufous-brown and 
thinner, more papery scales of E. angulatum. As in that species, the stipe bears 
broad, brown scales with jagged margins and these are continued on to the lower 
part of the midrib. The spores are considerably larger than in E. angulatum, 
agreeing in size with those of E. marginatum, from which, and from E. commutatum, 
it differs in its more strongly creeping rhizome, in its frond form and texture, and in its 
scale characters. 

Fee's description of E. stigmatolepis was based on a gathering " in herb, de Lessert., 
[sic] sub no. 159 ". The collector's name was evidently not stated on the label but 
the provenance is given as "in Indiis orientalibus, Neilgherries ". There is in the 
Cambridge University collection a sheet bearing a complete and well-preserved plant 
with three sterile and one fertile frond attached to a long-creeping rhizome 15 cm. 
long. No collector's name is inscribed on the label but this also reads " Ind. orient 
Neilgherries 159 ". The specimen agrees so exactly with Fee's description and plate 
that it is almost certainly an isotype. Wight 50 in herb. Kew is also an excellent 
match for Fee's illustration and the Cambridge plant. 

In Fee's description the fronds are said to be " membranaceis, papyraceis " and 
he does not include this species in the coriaceous-leaved group to which A . margina- 
tum, A . conforme and A . laurifolium are referred. At the same time the fronds are 
described as narrowly revolute, and re volute edges are commonly associated with the 
stiff-leaved species. I have seen no living plants of E. stigmatolepis but to judge from 
herbarium specimens the fronds are thin but stiff and are distinctly less coriaceous 
than those of the other marginate-leaved species. 

E. ballardianum Biswas is merely a redescription of E. stigmatolepis and inaccurate 
in the statement that the fronds lack a border. One of the gatherings which Biswas 
cites is labelled Acrostichum stigmatolepis and another has had the same name pen- 
cilled on the sheet, yet Biswas makes no reference to this species anywhere in his 
paper, though he could hardly have failed to note its description in the literature 
which he cites or that it was founded on plants from the identical area of his own 
supposed new species. 



88 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

3. Elaphoglossum beddomei Sledge, sp. nov. 

Elaphoglossum stigmatolepis sensu Bedd., Ferns S. Ind. : 67, t. 199 (1864) ; Handb. Ferns 

Brit. Ind. : 418 (1883) ; non T. Moore. 
Acrostichum stigmatolepis sensu Baker in Hook. & Baker, Synops. Fil., ed. 2 : 521 (1874) ; 

non F6e. 

Rhizoma repens, paleis linearihus brunneis marginibus integris 6-9 mm. longis basi 
I mm. latis vestitum, stipitibus crebrioribus. Frondes stipitatae, 20-35 cm. longae ; 
stipites frondium ster ilium plerumque breves et 3-6 cm. longi, vel, si longiores, 
dimidia parte longitudinis laminae breviores, juventutes paleaci paleis linearibus 
fimbriis marginalibus paucis ; lamina coriacea, anguste elliptica, 2-3 cm. lata, in 
apicem acutum diminuta, basi gradatim decrescens et in alas angustas aut cristas in 
stipite decurrens, marginibus angustis revolutis cartilagineis, infra paleis minutis 
stellatis vestita, supra in frondibus junioribus paleis similibus vestita, in vetustioribus 
glabrescens. Frondes fertiles eas steriles longitudine tota aequantes sed stipes 
quam lamina aequaliter longus ; lamina I-5-2-5 cm. lata. Sporae (36) 39-42 (45) x 
24-30 /I. 

South India: Anamallays, 1,050 m., " F.S.Ind. f. 199 Elaphoglossum stigmatolepis 
Kunze ", Beddome (BM, holotype). Palghat Hills, 1,500 m., Beddome (K). Nilgiri 
Hills, 2,100 m., Beddome (K). Same locality, Beddome 6 (K). Same locality 1,500 
m., 1870, Henderson (E). Sispara Ghat, Nilgiri District, Nov. 1883, Gamble 13418 
(K). 

South India (Nilgiri, Anamallay and Palghat Hills). 

Elaphoglossum beddomei differs from E. stigmatolepis in its longer and narrower, 
uniformly brown rhizome scales, the margins of which are entire or at most with an 
occasional marginal fimbriation. Its shortly stalked more coriaceous fronds are 
decurrent on the stipes, forming narrow wings or ridges often extending to the base, 
and the minute scales on the under surfaces of the fronds are stellately laciniated. 
The stipes bear linear scales markedly different from those of E. stigmatolepis, in 
which they are short and broad and are continued on to the midrib of the lower 
surface of the fronds. 

Beddome's illustration well portrays the habit of this species. In one of the three 
sterile fronds depicted the narrow wing extending downwards from the decurrent 
lamina to the base of the stipe is correctly shown. Sometimes the stipes are con- 
siderably longer than those illustrated, and then the decurrent wings merge into 
ridges on the sides of the stipes. The fringed scale represents one of those clothing 
the stipe, as the rhizome scales have margins which are mostly quite entire though a 
few fimbriations may sometimes be present. 

When Beddome first used the name E. stigmatolepis for this species he added a 
query both after the name and following the citation of Fee's figure. Baker's 
annotations on Beddome's specimens at Kew and his citation of Beddome's illustra- 
tion show that this was the plant which he had in mind when adding Acrostichum 
stigmatolepis to the second edition of the Synopsis Filicum and, following Baker, the 
query was dropped by Beddome in his Handbook. Two distinct species however occur 
in the Nilgiri and adjoining hills in South India and the true E. stigmatolepis is the 
plant which was subsequently distinguished by Biswas as E, ballardianum. He 






ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 89 

evidently accepted the present species, on the evidence of Beddome's illustration and 
the named sheets at Kew, as being E. stigmatolepis , though he makes no reference to 
this species anywhere in his paper but compares E. hallardianum with E. conforme, 
to which its relationship is certainly more remote. 

4. Elaphoglossum marginatum (Wall, ex Fee) T. Moore, Index Fil.: 8, ii 
(1857). 361 (1862). 

Acrostichum marginatum Wall., Numer. List : 2, n. 17 (1829), nom. nud. 
Acrostichum marginatum Wall ex F6e, M6m. Fam. Foug. 2:31 (1845), excl. syn. Blume. 
Acrostichum conforme sensu Hook., Sp. Fil. 5 : 198 (1864) pro parte. — Hook. & Baker, 

Synops. Fil. : 401 (1868) pro parte. 
Elaphoglossum conforme sensu Bedd., Handb. Ferns Brit. Ind. : 416 (1883) pro parte minore 
et excl. fig. 247 ; non J. Smith. ^ — Tagawa in Mem. Coll. Sci. Univ. Kyoto, Ser. B, 20 : 29 

(1951)- 
Elaphoglossum fuscopunctatum Christ in Bull. Herb. Boiss. 6 : 867 (1898) ; in Neue Denkschr. 
Schweiz. Ges. Naturw. 36 (i, 2) : 51 (1899). 

Rhizome stout, short-creeping, stipes close, clothed with dark brown, thick-walled, 
narrow, lanceolate, attenuate, shortly and irregularly fringed scales, 4-5 X i-i*5 mm. 
Fronds stalked, 20-50 cm., stipe of sterile frond up to half as long as lamina but 
often much less, deciduously scaly ; lamina elliptic, 2-4 (5) cm. broad, apex acute, 
base gradually narrowed and decurrent on the stipe, margins with a translucent 
cartilaginous border, lower surface with scattered minute, more or less stellately 
laciniated, brown scales, upper surface glabrous ; texture coriaceous. Fertile 
fronds on longer stalks than the sterile ones, lamina slightly narrower. Spores 
42-45 X 27-30 fi. 

Nepal: Sheopore, 1829, Wallich ly (K). Wallich 1^4 (BM). Sheopuri Lekh, 
Katmandu, 1,350 m., 19 Aug. 1954, Stainton, Sykes and Williams 6giy (BM). 
Between Torke and Okhaldunga, 1,950 m., 2 Nov. 1954, Zimmermann iggy (BM). 

SiKKiM : Darjeeling, 2,280 m., 9 Aug. 1875, C.B. Clarke 26gi3 (K). Same locality, 
1,800 m., Oct. 1881, Gamble g884 (K). Kohima, 1,800 m., 21 Oct. 1885, C. B. Clarke 
4ogyi (K). Neebay (?), 2,250 m., 16 Oct. 18—, C. B. Clarke 25337 (BM). Jeylep Rd., 
1,800 m., Nov. 1880, Gamble gg38 (K). Treutler gSS (K). 2,100 m., Sept. 1882, 
Levinge (K). Jerdon (K). 

Bhutan : Kancham, Punakha, 1,950 m., 28 Aug. 1914, Cooper 3058 (BM), 
Trashiyangsi Valley, Tobrang, on rock faces, dense jungle, 29 Aug. 1934, Ludlow & 
Sherriffgo2 (BM ; E). 

Assam : Sohra River, 1,500 m., 16 Oct. 1872, C. B. Clarke ig23g (BM ; K). Same 
locality, 1,500 m., 28 Nov. 1871, C. B. Clarke 14778 (K). Same locality, 1,350 m., 
15 Oct. 1872, C. B. Clarke 18826 (K). Vale of Rocks, Sept. 1886, C. B. Clarke 
45828 (K). Suruseem, Khasia, 1,200-1,800 m., 26 May 1850, Hooker & Thomson (K). 
Kalapanee, Khasia 1,650 m., 6 Aug. 1850 and 28 Oct. 1850, Hooker & Thomson (K). 
Assam, 1870-80, G. Mann (K). Ukhrul, Manipur State, on rocks in the forest, 
1,800 m., 12 Aug. 1948, Kingdon Ward I7g35 (BM). 

Himalayas eastwards from Nepal, Sikkim, Bhutan, Khasi Hills, Manipur, Yunnan, 
Formosa. 



go ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

Elaphoglossum marginatum differs from the other coriaceous-leaved species in its 
thicker, more shortly creeping rhizome with more crowded stipes. E. angulatum is 
readily distinguished by its long-creeping habit with well-spaced fronds and much 
broader, thinner and light brown rhizome scales. E. stigmatolepis also differs in its 
more rampant habit, its narrower fronds, which are more tapering above and below, 
and by the different form of the scales on the lower surface. Both Fee and Moore 
included Java in the distribution of E. marginatum through inclusion of the plants 
later distinguished by Mettenius as Acrostichum commutatum. From this E. mar- 
ginatum is best distinguished by its stouter rhizome and more clustered fronds and 
its darker coloured rhizome scales, which lack the long-attenuate fiexuose and curled 
apices and long marginal fimbriations of E. commutatum. 

Christ makes no mention of Acrostichum marginatum in his monograph of Elapho- 
glossum, but his E. fuscopunctatum described from Yunnan I consider the same as 
Wallich's plant. There are specimens at Kew and the British Museum of Henry 
9158, on which Christ's species was based, and these are, to me, indistinguishable 
from some Himalayan gatherings of E. marginatum (e.g. Hooker & Thomson, from 
Khasia in Herb. Kew ; Ludlow & Sherriff go2 from Bhutan in Herb. Edin.). The 
scales in several other collections of E. marginatum give a conspicuously fusco- 
punctate appearance to the undersides of the fronds. Tagawa (loc. cit.) reduced 
both to forms of E. conforme. 

5. Elaphoglossum commutatum (Mett.) Alderw. van Rosenb., Malayan Ferns, 

Suppl. I : 427 (1917). 

Acrostichum conforme sensu Blume, Envim. PL Jav. 2 : [275] (1828) (errore " aemulum " in 
p. loi) ; Fl. Jav., Fil. : 23, t. 5 (1829) ; non Swartz. — Hook., Sp. Fil. 5 : 198 (1864) 
pro parte. — Hook. & Baker, Synops. Fil. : 401 (1868) pro parte. 

Acrostichum marginatum F6e, Mem. Fam. Foug. 2 : 31 (1845) pro parte, incl. syn. Blume. 

Elaphoglossum laurifolium sensu T. Moore, Index Fil. : 359 (1862) pro parte, quoad specim. 
ex India, Ceylon, Java ; non T. Moore, op cit. : xvi (1857). 

Acrostichum laurifolium sensu Thw., Enum. PL Zeyl. : 380 (1864) ; non Thou. 

Acrostichum commutatum Mett. apud Kuhn in Ann. Mus. Bot. Lugd.-Bat. 4 : 292 (1869). 

Rhizome creeping, stipes rather close, clothed with narrow, brown scales with 
fimbriate margins and long-attenuate, fiexuose and twisted tips, 10 X 2 mm. Fronds 
stalked, 15-30 (50) cm., stipe of sterile fronds variable in length, up to half as long 
as the lamina but often much less, scaly when young with spreading, linear, fimbriate, 
dark brown scales and with smaller, appressed, stellately laciniated scales, but be- 
coming smooth with age ; lamina elliptic, 2-5 (6) cm. broad, apex acute, base 
gradually narrowed and decurrent on the stipe, margins with a translucent, cartila- 
ginous, revolute border, lower surface with scattered, minute, more or less stellately 
laciniated scales, upper surface with similar scales when young, becoming glabrous; 
texture coriaceous. Fertile fronds on stalks normally exceeding those of the sterile 
ones and sometimes the stipes longer than the sterile fronds, i'5-3'0 cm. broad. 
Spores 33-36 X 24-27 /I. 

South India : Bolampatty Hills, 1,500 m., Beddome (K). Nilgiri Hills, 1,200 m., 
Nov. 1870, Henderson (E). 



ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 91 

Ceylon: Thwaites C. P. i 310 {BM ; CGE ; E; K; PDA). Hoolankande Peak, 
Matale, 10 Jan. 1862, Brodie (E). Nuwara Eliya, Freeman 368 A (BM). Kabot's Gap, 
Rangala, 8 Sept. 1927, Alston 2375 (PDA). Knuckles, 1881, no collector's name 
(PDA). Ramboda Pass, 1,650 m., epiphytic on tree trunk by stream in jungle, 
17 Dec. 1950, Sledge 646 (BM). Hoolankande, 1,350 m., 20 Jan. 1954, Sledge loio 
(BM). Jungle above Le Vallon Estate, 1,500 m., 9 Feb. 1954, Sledge 1116 (BM). 
Parawella Falls, Kandapola, near Nuwara Eliya, 1,425 m., 19 March 1954, Sledge 
I32g (BM). Robinson 20 (K). Central Prov., Naylor Beckett (K). 1871, Randall 
in Herb. R. W. Rawson 3285 (BM). R. W. Rawson I02y-i802 (BM). Henderson (E). 
Anderson (E). 

South India (Nilgiri Hills), Ceylon, Java. 

Elaphoglossum commutatum is intermediate in habit between E. marginatum and 
E. angulatum, with less crowded stipes than the former but with shorter internodes 
than the latter, in which the fronds are 1-2 cm. or more apart. It differs from 
E. marginatum in its rhizome scales as described under that species, and in its smaller 
spores. From E. angulatum, apart from its much less widely creeping rhizome, it 
differs in its markedly narrower, darker brown, finely acuminate and strongly 
fimbriate rhizome scales. The stipes bear similar linear dark-coloured scales inter- 
mixed with small appressed stellate scales like those of the under surface of the 
blades, which spread down the stipes, though in old fronds the stipes may be practi- 
cally naked. In E. angulatum the broad, pale brown scales which clothe the rhizome 
spread up the stipe and are continued on the lower part of the midrib of the lamina, 
becoming narrower and more laciniated upwards and gradually merging into the 
minute stellate scales of the under surface. Stellate scales are absent on the stipes of 
E. angulatum. 

Mettenius based his description (loc. cit.) on plants of Blume's from Java and on 
Thwaites C.P. 1310 from Ceylon. His references to Thwaites's numbered collection 
and to the description and figure of Blume's A. conforme and his own accurate 
description of the rhizome scales fix the identity of the species. Van Alder wereldt 
van Rosenburg validly transferred the specific epithet to Elaphoglossum though the 
plants he had in mind were doubtfully the same as those of Mettenius. The species 
has been treated in most works as a synonym of E. laurifolium (Thou.) Moore, 
which was described from Tristan da Cunha. Christensen (Res. Norw. Sci. Exped. 
Tristan da Cunha, 1937-38, 6 : 19 (1940)) expressed the view that E. laurifolium was 
endemic to Tristan da Cunha. Through the kindness of Mr. James Dickson I have 
been able to examine good specimens of E. laurifolium collected by him on Tristan da 
Cunha in 1962. These and other gatherings from Tristan and from Gough Island 
stand well apart from Ceylon and South Indian plants in their wide-creeping habit 
with long internodes, their dark brown-black rhizome scales, which lack the fim- 
briated edges and crisped apices of E. commutatum, and their larger spores (45-48 X 
30-33/^). 

E. commutatum stands between E. laurifolium and E. callifolium, the latter being 
a larger plant with bigger rhizome scales (up to 2 cm. long), which have entire 
margins. Backer and Posthumus (Varenfiora voor Java: 251 (1939)) maintained 
E. commutatum (sub. nom. E. laurifolium) and E. callifolium as distinct species, 



92 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

though they express the view that the former may be no more than a form of the 
latter. I have no doubt as to Ceylon and South Indian plants being distinct from 
E. callifolium but I am much less sure if plants from Java and elsewhere in Indonesia 
which have been called E. laurifolium are identical with those from India and Ceylon. 
Specimens so named from Java, Sumatra, Borneo and Indo-China have rhizome 
scales the margins of which are less irregular and they lack the frizzed tips of those of 
E. commutatum. They approach E. callifolium in these respects though the scales 
are narrower and shorter. 

I suspect E. permutatum Alderw. van Rosenb. (Bull. Jard. Bot. Buitenz., ser. 2, 
i6 : 13 (1914)) to be synonymous with E. commutatum. There is an isotype {Matthew 
6g6 from Mt. Sago, Sumatra) at the British Museum and this has rhizome scales with 
fimbriate margins and crisped apices although the description states that the scales 
are " integerrimis ". In its other characters the cotype does not appear to me to be 
significantly different from E. commutatum. 

6. Elaphoglossum stelligerum (Wall, ex Baker) T. Moore ex Alston & Bonner 
in CandoUea 15 : 216 (1956). 

Acrostichum neriifoHum Wall., Numer. List : 2, n. 16 (1829), nom. nud. 
Acrostichum stelligerum Wall., op. cit. : 65, n. 2167 (1830), nom. nud. 
Elaphoglossum stelligerum T. Moore, Index Fil. : 15 (1857), 368 (1862), nom. nud. 
Acrostichum viscosum sensu Hook., Sp. Fil. 5 : 220 (1864) pro parte ; non Swartz. 
Elaphoglossum viscosum sensu Bedd., Ferns. S. Ind. : 67, t. 196 (1864) ; Handb. Ferns 

Brit. Ind. : 420 (1883) ; non J. Smith. 
Acrostichum stelligerum Wall, ex Baker in Hook. & Baker, Synops. Fil, ed. 2 : 521 (1874). 
Acrostichum. yunnanense Baker in Bull. Misc. Inf. Kew 1898 : 233 (1898). 
Elaphoglossum yunnanense (Baker) C. Chr. in Contrib. U.S. Nat. Herb. 26 : 327 (1931). 

Rhizome creeping with stipes close together, clothed with linear, attenuate, dark 
castaneous, glossy scales, 3-5 X 0-5 mm., with occasional fimbriations in their lower 
parts, elsewhere with entire or obscurely dentate margins. Fronds stalked, 20-40 
cm. long, stipes of sterile fronds densely scaly with a mixture of stellate scales with 
slender rays and lanceolate scales with long-fimbriate margins ; lamina narrowly 
elliptic 1-2-5 cm. broad, apex acute, base gradually narrowed and decurrent on the 
stipe, lower surface with abundant, loose, reddish-brown, stellate scales with long rays 
which diverge from the leaf surface, the midrib with similar scales intermixed, especially 
in the lower part, with broader-centred ones with long-fimbriate margins ; upper 
surface stellate-pubescent when young, becoming more or less smooth with age, the 
scales paler than those on the lower surface ; texture thin but firm. Fertile fronds 
on longer stipes than the sterile ones, lamina narrower, to i cm. broad. Spores 
42-45 X 30 /I. 

Nepal: 1820, Wallich 16 (BM ; E; K). Wallich ex coll. Hooker f & Thomson g6 
(E). Panapa, 1,546 m., 6 Sept. 1954, Zimmermann 1036 (BM). Between Maneb- 
hanjyang and the Sun Kosi, 1,500 m., 5 Nov. 1954, Zimmermann 2058 B (BM). 

SiKKiM : 600 m., Beddome (BM). 2,400 m., 1868, Henderson (E). 1871, Levinge 
(K). Darjeeling, 1871, Hope (E). Same locality, coll. Lmwg^, ex Herb. Hope (E). 
Same locality, Oct. 1873, C. B, Clarke 21^86 (K). Goke, 17 July 1880, Levinge (E). 



ELAFHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 93 

Bhutan : Towards Kolepani, 3 Nov. 1835, Griffith (K). 

Assam : Khasia, 1,350 m., Hooker & Thomson (E ; K). Same locality, 1,200 m., 
16 Sept. 1850, Hooker & Thomson 2285 (K). Same locality, 1,200 m., 16 Nov. 1871, 
C. B. Clarke 146J8 (K). Jowye, Jainteas, 1,200 m., 19 Oct. 1867, C. B. Clarke ^gjS 
(K). Mariao, 1,200 m., 7 Nov. 1871, C. B. Clarke i536g (BM ; E). Bogapani, 
Khasia, 1,050 m., 29 Oct. 1871, C. B. Clarke i6;^6y (K). Same locality, 1,200 m., 
July 1890, G. Mann (E). 

South India : Pulney Mts., Sept. 1836, Wight $1 (K). Same locality, 1,800 m., 
Beddome (BM ; K). Periya Shola, Pulneys, 21 Apr. 1898, Bourne 4gyi (K). Kodai- 
kanal, Pulney Hills, 1,960 m.. May 1937, Aroticasamy 7 (BM). Nilgiri, i860, 
Beddome 4 (K). Anamallays, 1,350 m., " F.S.Ind. f. 196 ", Beddome (BM). Same 
locality, 1864, Henderson (E). Shevaroy Hills 1,200 m., 1898, coll. /. W. Furrell 66 
ex Herb. Hope (E). Massif du Shevaroy, 10 Oct. 1939, Faucheux (BM). Honey Rock 
stream, Yercaud, Shevaroy Hills, 1,100 m., 25 Oct. 1962, Ghatak Gig^A (BM). 

Yunnan, Indo-China. 

Elaphoglossum stelligerum is readily distinguished from all other Indian species 
by its loose, rufous-brown, stellate scales with long rays which are not appressed to 
the surfaces of the stipes or fronds. The linear, dark-coloured, somewhat glossy 
rhizome scales are also markedly different from those of other species. It is closely 
related to E. petiolatum (Swartz) Urban, described from Jamaica and formerly 
credited with a pantropic distribution. 

I see no significant difference between E. stelligerum and Baker's Acrostichum 
yunnanense. Baker's description states " ad ^. stigmatokpidem, Fee, magis accedit; 
frondibus linearibus ad marginem paleis parvis ciliatis et paleis facialibus profunde 
stellatim fissis differt " ; but the comparison with E. beddomei — A. stigmatolepis 
sensu Baker being E. beddomei mihi — is incomprehensible since the plant Baker had 
before him is manifestly far closer to, and is indeed in my opinion specifically identical 
with, E. stelligerum. Ching evidently dismissed it as a valid species since there is a 
second sheet of Henry 103 10, the type gathering of A . yunnanense, in the same folder 
at Kew which he has labelled E. petiolatum (Sw.) Urban. Christensen upheld 
E. yunnanense (in Contrib. U.S. Nat. Herb. 26 : 327 (1931)) but his comments on the 
distinction between it and E. petiolatum are equally applicable to E. stelligerum, 
which he did not distinguish. Nor do I agree with Holttum (Fl. Malaya 2 : 455 
(1954)) that Malayan specimens agree better with the type of E. yunnanense than 
with South Indian specimens. They are more robust than either Henry's Yunnan 
collection or Indian gatherings. 

E. stelligerum is very closely related to E. blumeanum (Fee) J. Sm. from Malaysia 
and E. salicifolium (Willd. ex Kaulf.) Alston from the Mascarene Islands (type from 
Reunion) and Africa, all being formerly included in E. petiolatum Swartz, which was 
described from Jamaica. Malaysian plants are larger than those from India and 
the scales on the two surfaces of the sterile fronds are dissimilar. Indian and 
Mascarene plants have the scales on the two surfaces alike and differences between 
them are so slight that they probably represent no more than geographical variants. 
The names are retained however pending a much-needed revision of the whole 
E. petiolatum complex. 



94 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

7. Elaphoglossum nilgiricum Krajina ex Sledge, sp. nov. 

Elaphoglossum squamosum sensu Bedd., Ferns S. Ind. : 67, t. 197 (1864) ; Handb. Ferns 
Brit. Ind. : 420, fig. 251 (1883) excl. specim. ex Ceylon ; non J. Smith. 

Rhizoma breviter repens, paleis anguste lanceolatis vel linearis, pectinato-dentatis, 
fusco-castaneis, marginibus nigris, nitidis, vestitum. Frondes caespitosae, breviter 
stipitatae, 5-25 (40) cm. longae ; stipites paleis ovatis vel late ovatis, pallido-ferru- 
gineis in margine dentibus setiformibus praelongis, densissime obtecti ; lamina 
anguste oblonga vel longitudo-ligulata, 1-2 cm. lata, apice obtusa, in basi sensim 
attenuata, in marginibus et utraque superficie copiose squamosa, infra paleis brun- 
neis, mollibus, lanceolatis ad late ovatis, margine dentibus setiformibus fimbriata, 
confertis imbricatis occultata, supra paleis similibus minus confertis, pallidioribus, 
aliquibus paleis in ambitu plus-minusque orbiculatis sed in margine radiis tenuibus 
longis instructis ; textura crassiter herbacea. Forma frondium fertilium ut in 
sterilibus. 

South India : Sisparah Ghat, Nilgiris, 1,500 m., " type of tab 197. FSL", 
Beddome (K, holotype). Same locality, 1,800 m., Nov. 1883, Gamble 13468 (K), 
Same locality, Miss Cockburn 81 (BM). Nilgiri, i860, Beddome 7 (K). Pykara Falls, 
Nilgiri^; District, 1,800 m., June 1883, Gamble i20og (K). Nediwattan, Nilgiri Hills, 
1,800 m.. May 1866, Gamble ex Herb. Blanford (E). Nilgiri Hills, Henderson (E). 

Endemic to Nilgiri Hills. 

The type of Elaphoglossum hirtum (Swartz) C. Chr. was described from Jamaica. 
This name, or the synonym E. squamosum (Swartz) Urban, was formerly used in a 
collective sense to cover plants ranging from tropical America eastwards to southern 
India and Ceylon, which are the only regions in Asia whence plants have been found. 
Several species based on collections from Africa, Madagascar, and the Mascarene 
Islands have since been given independent rank. As South Indian and Ceylon plants 
not only differ from each other but cannot be matched satisfactorily with any 
Jamaican, African or Mascarene gatherings which I have examined, I have adopted 
the unpublished names affixed to the Kew sheets by Krajina in 1937. 

Specimens from the Nilgiri Hills are very uniform and they differ consistently 
from Ceylon plants in the scales of the stipes being markedly broader with their 
margins and setiform teeth rarely dark-coloured as in £. ceylanicum, save at the base 
of the stipe adjacent to the rhizome. They also differ in their much more densely 
paleate fronds, the lower surfaces of which are hidden beneath the closely imbricating 
pale-brown scales. 

Beddome's illustration (Ferns S. Ind. : t. 197 (1864)), is a fair representation of 
the species except that the stipes are unaccountably portrayed as smooth and glabrous 
whereas his description states that they are " very scaly " and both in his own and 
all other gatherings they are invariably densely paleate with patent scales. The 
isolated, long, falcate and apparently glabrous frond depicted on the left-hand side 
of the plate is meaningless and, not surprisingly, is omitted from the reproduction 
of this plate in the Handbook. 



ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 95 

8. Elaphoglossum ceylanicum Krajina ex Sledge, sp. nov. 

Rhizoma breviter repens, paleis linearibus pectinato-dentatis fusco-castaneis vel 
nigris vestitum. Frondes caespitosae, breviter stipitatae, 5-25 cm. longae ; stipites 
paleis anguste lanceolatis, attenuatis, acutis, ferrugineis, in margine dentibus 
setiformis praelongis, plerumque marginibus dentibusque fuscis, obtecti ; lamina 
oblongo-elliptica, 1-2 cm. lata, in apice acuta vel acuminata, in basi sensim attenuata, 
in marginibus et utraque superficie paleis tenuibus brunneis ; supra pallidioribus 
vestita, sed infra, praeter in juventute, non paleis occultata. 

Ceylon : Gardner 1164 (BM ; CGE on trees in dark forests at Nuwara Eliya, 
Sept. 1844; K, holotype). Thwaites, C-P- 3292 {BM ; CGE; E; PDA). Central 
Province, on trees and rocks, Naylor Beckett 401 (K). Peacock Hill, Pussilawa, 
i8yo, Randall in Herb. R. W. Rawson 3284 (BM). Pundaloya, 1,500 m.. Freeman 
372 A, 374A, 37 5B (BM). Nuwara Eliya, Freeman 373B (BM). Walker (K). 
Robinson 22 (K). Rocks by stream at Lonach near Norton Bridge, Central Province, 
900 m., 13 Dec. 1950, Sledge 398 (BM). Ramboda, on mossy rock by stream in jungle, 
1,650 m., 17 Dec. 1950, Sledge 648 (BM). Kuda Oya, Ramboda, on rocks in shade 
by stream through jungle, 1,725 m., 28 Dec. 1950, Sledge 769 (BM). 

Endemic. 

The distinctions between this species and E. nilgiricum are referred to under the 
previous species. The fully developed fronds are acute at the apex and green in 
colour since neither surface is covered by a continuous coating of scales as is the lower 
surface of the frond in E. nilgiricum. Young fronds however are very scaly. In 
both species the lamina scales vary in shape, the body of the scales being mostly 
broadly to narrowly ovate in outline but some are more or less orbicular. All have 
the margins fringed by long setiform teeth. 

E. ceylanicum seems to come closest to E. deckenii var. rufidulum (Willd. ex Kuhn) 
Tardieu-Blot (Not. Syst. 15 : 430, t. 4 f. 6-10 (1959) =E. hirtum var. rufidulum 
(Willd. ex Kuhn) C. Chr. in Dansk. Bot. Ark. 7 : 170 (1932)) from Madagascar and 
Reunion. In that species, however, the fronds are borne on long stipes and the 
shape of the rhizome scales and their marginal teeth are different. 

Plants from Madeira and the Azores have jet black scales on the stipes and under 
surface of the costae, intermixed with ferrugineous ones with and without thickened 
and darkened margins. Such scales are not present in E. nilgiricum or E. ceylanicum. 
Specimens from Madeira and the Azores are closest to E. splendens (Bory) Brackenr. 
from Reunion and Mauritius but are better treated as a separate species.^ 

^ Elaphoglossum paleaceum (Hook. & Grev.) Sledge, comb. nov. 

Acrostichum paleaceum Hook. & Grev., Ic. Fil. 2 : t. 235, Alph. Index et Syst. Index (1831) (errore 

" vestitum " in expl. tab.). 

It would seem that Hooker and Greville realized during the preparation of the Icones Filicum that 

Lowe's manuscript name Acrostichum vestitum, was a later homonym. Its appearance in the text 

accompanying the plate, in spite of the fact that A . paleaceum is used on the plate and in the indexes, is 

to be regarded as an error. 



96 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 

9. Elaphoglossum spatulatum (Bory) T. Moore, Index Fil. : 14 (1857). — 
Bedd., Ferns S. Ind. : 71, t. 209 (err. 213) (1864). 

Acrostichum spatulatum Bory, Voy. Mers d'Afriq. i : 363, t. 20 f. i (1804). — F6e, M6m. Fam. 

Foug. 2 : 51, t. 14 f. 3 (1845). 
Acrostichum acutum F6e ex Kuhn in Linnaea 36 : 44 (1869). 

Rhizome creeping with stipes rather close together, clothed at the apex with 
filiform, rufous scales. Sterile fronds up to 10 cm. long, stipes about equalling the 
lamina in length and covered with patent, reddish, setaceous scales ; lamina lanceolate 
to rhomboid-lanceolate, 1-1-5 cm. broad, base narrowed into the stipe, apex usually 
acute sometimes obtuse, margins and both surfaces covered with hair-like scales ; 
texture rather thickly herbaceous. Fertile fronds smaller than the sterile ones and 
on much longer stipes, lamina broadly oval or almost orbicular, conduplicately 
folded when young and often splitting at the apex when mature. 

Ceylon : Thwaites C.P. 989 (BM ; E ; K ; PDA). Adam's Peak, Moon (BM). 
Nuwara Eliya, Rawson W. Rawson 1041 (BM). Same locality, Freeman 376 A, 
377B (BM). Sita Eliya, Hakgala, March 1885 (PDA). Hawa Eliya, 25 Aug. 1926, j 
/. M. de Silva (PDA). Kuda Oya, Ramboda Pass, 1,700 m., in shady places on rocks 
by stream in jungle, 28 Dec. 1950, Sledge yji (BM). 1,500-1,800 m. " Type of tab 
209 FSI ". Thwaites in Herh. Beddome (K). 1870, Thwaites in Herb. Henderson (E). 
On wet rocks at about 1,800 m., Hutchison (E.). Wall (E). Robinson 23 (K). 
Gardner (K). 

Africa (N. & S. Rhodesia, Transvaal, Natal, Cape), Madagascar, Reunion. 

A very distinct species, much smaller than all other Indian and Ceylon species. , 
Closely related plants, which have often been treated as belonging to the same species, I 
occur in tropical America and Tristan da Cunha ; but Christensen (Res. Norw. Sci. 
Exped. Tristan da Cunha, 1937-38, i (6) : 20 (1940)) considered the species on that 
island, E. obtusatum (Carm.) C. Chr., to be " certainly quite different from E. spathula- 
tum " . 

Madame Tardieu-Blot (Not. Syst. 15 : 432 (1959) ; in Humbert, Fl. Madag., 
Polypod. 2 : 49 (i960)) has erroneously transferred the Ceylon records of E. spatulatum 
to E. schizoUpis (Baker) Christ. The type of the latter — from Madagascar — is at 
Kew and is quite unlike any Ceylon plant. 








M^'^^-'ii^'.'ifK'r Si^^ ■'«.te'r*^9'^.v \:;v ''>'-'-;v':^: ;:-'^i'''!/.'i. ?;■■'■-•":•' ..,-.''' ; ': 'v; '.V' ';! ■-^^■ - v'^ ' •■ '."'.' 

BW'f'-i*f5-;y;-^' V 'v^.- r*'.,. ■;''i''*^ ."7,,, -. ; , ,V,^-'-: v. , ' - -v-lV ,'•* ,\,' >"■■,>'■ j^'*. •"''..■■ I'V' *^- ■', ', * v ' ,. ' .' ■• ."■■■■ /> -' '^^ 

?^r-rJ^■v-■■tv!V'*t' v^ ■■■■^'V-''-'' '^ ■''^ ■ •■"■ v"-V- '■ ^^^^ ■'..■ vV •''■'' " <'•■;■■ -■'"■■. ': • ; •-■ .' '■■'-'; ' ■■■" 

^l¥:ii^r l!■^ -^^^ri^^s^ ■'; -■'"'■--"■■:'■■•''•.■■"' ■■'■^ .\ ■'/ ■ v.-.';:'^' '.• ■ ' ;/'■■■,; 

^',y«^':'^^'^^.j'*jv;;^,;.:it,„s.>;/v,^ x,-,>';x ,,^:^;,;- v^ -,;■■:;■;■•-■;;:.;■•, .".i ;■':;:''' ■::-^ :-rA. • 

K<!^PW;^'.'V;:' -■«»,V-''>-Vvj> ■ .■'■"■','v'''-. r'^^i'-rv' -■■'■■■ - •■ t'-'.'>-'. -i': '.r' 'V '' ■ ■■:■•..'■■ ■' '. ■'-: ■ "^-^ .:■■'<•■ •.^i ::>i ' ;• ■' 

i^^illif ^i:|j||?:: •:^JS^ ^' r-(^;>^-S J:>|>^/^i" ■' ■ ;' ■ ^^^' -?'': f'^:^ ■^^i'^v '^t! '>^' 



:^*;i 







'■*.--"'^ ■■;■:■;. r: 



FUNGI OF RECENT NEPAL 
EXPEDITIONS 



FRANCES L. BALFOUR-BROWNE 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 3 

LONDON: 1968 



FUNGI OF RECENT NEPAL EXPEDITIONS 




BY 

FRANCES L. BALFOUR-BROWNE, 



/^ 



Pp. 97-141 ; 4 Text-figures 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 3 

LONDON: 1968 



THE BULLETIN OF THE BRITISH MUSEUM 

(NATURAL HISTORY), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 3 of the Botany 
series. 



World List abbreviation: 
Bull. Br. Mus. nat. Hist. {Bot. 



Trustees of the British Museum (Natural History) 1968 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 24 June, 1968 Price Nineteen Shillings 



FUNGI OF RECENT NEPAL EXPEDITIONS 

By FRANCES L. BALFOUR-BROWNE 

The fungi recorded below were collected during four general botanical expeditions 
made in the Nepal highlands by (i) J. D. A. Stainton, W. R. Sykes and L. H. J. 
WiUiams in 1954, (2) J. D. A. Stainton in 1956, (3) A. H. Norkett in 1961-62 and (4) 
J. D. A. Stainton in 1962. Expeditions (i) and (4) were in Central Nepal while 
the other two were mainly in Eastern Nepal. 

The collections were made at altitudes between 300 m. and 5,000 m. At the 
highest levels were Conifer and Rhododendron forests with Evergreen Oaks coming 
in from below, and the lower levels were characterized by rice terraces, bamboos, 
ferns and Castanopsis forest. Sheals & Inglis (1965) give details of the local geog- 
raphy and the nature of the terrain encountered in Expedition (3). 

In this account references have been restricted to the authorities of the names 
used and to the more significant or well-known synonyms. For additional references 
to most of the species the revised (i960) edition of Butler & Bisby's indispensable 
Fungi of India by Vasudeva should be consulted. 

In some groups the delimitation of genera is very controversial and consequently 
the classification and the names employed are very much a personal matter of 
opinion with very little agreement amongst workers. This is particularly so for 
the Polyporaceae. During the last twenty years new systems of classification for 
this group have been published by Pilat (1936-42, Europe), Cunningham (1947, 
1947-50, 1965, New Zealand), Corner (1932, 1953, general), Overholts (1953, America), 
Imazeki (1943, Japan), Donk (1933, i960, general), Singer (1962, general), Bondart- 
zev (1953, Russia), Nobles (1958, Canada), Pinto-Lopez (1952, Portugal), Kotlaba 
& Pouzar (1957, Europe), Teixeira (1962, Brazil). These systems have been based 
on a variety of criteria: general anatomical and morphological structure, hymenial 
structure, hyphal structure, biosystematics, hyphal thickening and the presence or 
absence of clamps, and physiological characters. Significant as are these characters, 
nevertheless owing to the difference in emphasis placed upon them by different 
authors, considerable difficulty arises in attempting to derive a stable or consistent 
nomenclature. Here, therefore, the well-known, old or mainly Friesian subdivisions 
have been used for the Polyporaceae and the less conservative names included in 
the synonymy. 

Recently suggested, but still tentative, relationships of certain agarics (so hitherto 
regarded) with polypores rather than with other agarics, or vice versa, are also not 
taken up here. The second edition of R. Singer's The Agaricales in Modern Taxo- 
nomy (1962) should be referred to for new ideas on the classification of this group. 

As regards the Clavariaceae, these have been named by Dr. E. J. H. Corner of 
Cambridge, and each identification for which he is responsible is indicated by his 
initials in parentheses. 

In mountainous countries such as Nepal with an annual rainfall in some areas 
of 500 cm., the difficulty is to get the plants dry. The tendency is therefore to 
press them too enthusiastically, with the consequence that some of the agarics and 

BOT. 4, 3 7 



loo FUNGI OF RECENT NEPAL EXPEDITIONS 

Clavariae were tissue-paper thin and their hyphal structure indiscernible. To off-set 
this it would have been valuable to have had some of the soft and fleshy fungi 
preserved in fluid, and full notes as to colour, texture and shape when fresh are 
always much desired. For the rest, the material was in good condition and together 
represents the largest collection so far from this difficult and until recently almost 
inaccessible region: i6o species excluding some immature and over-ripe specimens 
which so far have resisted identification. Previous records consist of about two 
dozen species collected by Sir Joseph D. Hooker over a century ago and described 
by Berkeley (1850, 1851, 1852, 1854) ^-^^d about 70 species collected by Polunin, 
Sykes and Williams in 1952 and reported in a previous number of this journal 
(Balfour-Browne, 1955). 

All the specimens cited in this paper are in the herbarium of the British Museum 
(Natural History). 

PHYCOMYCETES 

ALBUGINACEAE 

Albugo bliti (Biv.) Kuntze, Revis. Gen. PL 2 : 658 (1891). — Best. Biga in Sydowia 
9 : 347 (1955)- 

Uredo bliti Biv., Stirp. Rar. Sicil. 3:11 (1815). 

Cystopus bliti (Biv.) De Bary in Ann. Sci. Nat., Ser. 4, Bot. 20 : 131, tab. 13 figs. 13-15 

(1863). 
Caeoma amaranthi Schwein. in Trans. Amer. Phil. Soc, New Ser., 4 : 292 (1832). 
Cystopus amaranthi (Schwein.) Berk, in Grevillea 3 : 58 (1874). 

Nepal: Bhurungdi Khola, 1,600 m., on Amaranthus lividus L., 20th May, 1954, 
Stainton, Sykes & Williams 5342. 
Distribution: Worldwide. 
For a recent review of the genus Albugo consult Bestagno Biba (tom. cit. : 339-58). 

PERONOSPORACEAE 

Sclerospora graminicola (Sacc.) Schroet. in Cohn, Krypt.-Fl. Schles. 3 (i) : 236 
(1886). 

Protomyces graminicola Sacc. in Nuov. Giorn. Bot. Ital. 8 : 172 (1876). 

Nepal: Bongakhani, 2,130 m., on grass, 22 Aug. 1954, Stainton, Sykes S" Williams 

3943- 

Distribution: America, Europe, Africa, India, Australia. 

PYTHIACEAE 

Phvtophthora infestans (Mont.) De Bary in Journ. Roy. Agric. Soc, Ser. 2, 1.2 :^ 
2^0 (1876). 

Sotrytis infestans Mont, in I/Institut, Sect, i, 13 : 313 (1845). 

Nepal: Murigurja Gad, 2,500 m., on potato, 27 July 1954, Stainton, Sykes 
Williams 3654. 
Distribution: Worldwide. 



FUN(;i OF RECENT NEPAL EXPEDITIONS loi 

ASCOMYCETES 

MORCHELLACEAE 

MoRCHELLA ELATA Fries, Syst. Mycol. 2 : 8 (1822). 

Nepal: Siklis, north of Pokhara, 3,000 m., on rotten tree trunk, 21 Apr. 1954, 
Stainton, Sykes & Williams 4956. 

Distribution: America, Europe, India, China, Japan, Austraha. 

HELVELLACEAE 

Helvella crispa Fries, Syst. Mycol. 2 : 14 (1822). 

Nepal: Lete, 2,600 m., beneath conifers, 27 Aug. 1954, Stainton, Sykes & Williams 

7501. 

Distribution: Worldwide; previously recorded from Nepal in 1955. 

humariaceae 

Aleuria aurantia (Fries) Fuckel in Jahrb. Nass. Ver. Naturk. 23-24 : 325 (1870). 

Peziza aurantia Fries, Syst. MycoL 2 : 49 (1822). 

Nepal: Ghar Khola, 2,130 m., 14 June 1954, Stainton, Sykes & Williams 5759. 
Distribution : Worldwide. 

GEOGLOSSACEAE 

Geoglossum affine (Durand) Sacc. & Trav. in Sacc, Syll. Fung. 19 : 756 (1910). — 
Maas Geest. in Persoonia 4 : 23 (1965). 

■ Gloeoglossum affine Durand in Ann. Mycol 6 : 420 (1908). 

" Nepal: Annapurna Himal, Seti Khola, 3,830 m., on shady banks, 28 July 1954, 
Stainton, Sykes & Williams 6540. South of Gurjakhani, 4,000 m., among dwarf 
Rhododendron, 17 Aug. 1954, Stainton, Sykes & Williams sSyg. 
Distribution : United States, Himalayas. 
These specimens were examined and identified by Dr. Maas Geesteranus. 

Mitrula ROSEA Lloyd, Mycol. Not. 61 : 885, t. 129 fig. 1529 (1919). 

Nepal: Gosainkund, Malemchi, 2,800 m., on bare earth, 30 May 1962, Stainton 
37S9. 

Distribution: India. 

Pale rose coloured ascophores about i cm. tall with smooth or contorted caps. 

sclerotiniaceae 

RuTSTROEMiA sp. cf. FiRMA (Fries) Karst. in Bidr. Kann. Finl. Natur & Folk 19 : 
108 (1871). 

Peziza firma Fries, Syst. Mycol. 2 : 117 (1822). 

Ciboria firma (Fries) Fuckel, Symb. Mycol. : 312 (1869). 

Phialea firma (Fries) Gill., Champ. Fr., Discom, : loi, t. 74 fig. 2 (1883). 



I02 FUNGI OF RECENT NEPAL EXPEDITIONS 

Nepal: Dhankuta Province, near Mahe, 1,300 m., on dead twig, 20 Sept. 1961, 
Norkett 5175 B. 

Distribution (of R. firma) : Worldwide. 

The fungus agrees well with this species except that its spores are somewhat 
short, 9-11 X 4-5 /*. The material, however, is barely ripe. 

HYALOSCYPHACEAE 

Perrotia malemchiensis Balfour-Browne, sp. nov. (Fig. i.) 

Apothecia sparsa, superficialia, sessilia, carnosa, uda pallide brunnea, sicca cinna- 
barina, extus villosa, 500-1,000 /i diam. ; setae cylindraceae ad basim i cohaerentes 
apice liberae et acutae, septatae, minute granulosae, pallide ochraceae ; asci clavati, 
recti vel curvuli, octospori, 80-110 X 9-10 /*; sporae clavatae vel fusoideae, hyalinae, 
5-7 septatae, 20-30 x 4-5 /i; paraphyses iiliformes, septatae, hyalinae. 

Ad ligna et cortices arborum frondosarum. 




«, 



I 



Fig. I. Perrotia malemchiensis Balfour-Browne. a, asci containing ascospores; 
B, paraphyses; c, ascospores; d, external hair; e, apothecium. Stainton 3y68. 



FUNGI OF RECENT NEPAL EXPEDITIONS 



103 



Nepal: Gosainkund, Malemchi, 2,400 on a dead tree, 29 May 1962, Stainton, 
3y68 (holotype). 

This differs from related species not only in colour and measurements but in spore 
septation: P. fusca Miill. & Dennis, i-septate; P. lutea (Phill.) Dennis, up to 31- 
septate; P. himalayensis Miill. & Dennis, 3-septate. 

DERMATEACEAE 

Mollisia dhankutae Balfour-Browne, sp. nov. (Fig. 2.) 

Apothecia superficialia, sessilia, usque ad 1-5 mm. diam., disco sordide flavido- 
albo; excipulo fusco-brunneo, pseudoparenchymatico ; asci cylindraceo-clavati, 
octospori, poro jodo tincto, 80-90 X9-io/«; ascosporae biseriatae, elongato- 




FiG. 2. Mollisia dhankutae Balfour-Browne. a, asci and paraphyses; 
B, ascospores; c, apothecia. Norkett 7751 A. 



I04 FUNGI OF RECENT NEPAL EXPEDITIONS 

fusoideae, uniseptatae, hyalinae, 16-20 X 3-5-4 /*; paraphyses filiformes, septatae, 
2 /I, supra usque 3-5 [i. 
In ramis siccis Bambusae. 

Nepal: Dhankuta Province, Taplejung district, Sanghu, 1,750 m., on dead 
bamboo stump, 19 Nov. 1961, Norkett yy^i A (holotype). Sanghu, 2,060 m., 9 Nov. 
1 96 1, Norkett 7349. 

Although very close to Mollisia caesia var. andina Dennis (Kew Bull, 14 : 441 
(i960)) from Venezuela, it differs in the absence of a whitish margin to the receptacle, 
in the slightly larger and more acutely pointed spores and in the different host. It 
differs also from Cenangella bambusicola Rick (Broteria 5 : 37 (1906)), which was 
described from living bamboo in South America, in that the paraphyses do not form 
an epithecium and the apothecia are not at first white, but externally very dark 
brown with an off-white disc. 

OSTROPACEAE 

ViBRissEA TRUNCORUM Fries, Syst. Mycol. 2 : 31 (1822). 

Nepal: Rambrong, Lamjung Himal, 4,500 m., on dead roots of Rhododendron, 
10 July 1954, Stainton, Sykes & Williams 626g. 

Distribution : America, Europe. Apparently not previously recorded from Nepal 
or neighbouring countries, possibly because it is easily overlooked, being small and 
generally found on roots, frequently submerged in water, or on debris. 

Apothecia scattered or in groups, with orange discs, 3-5 mm. diam., seated on 
pale stalks, blackish below, 15 X 2-3 mm. Asci 250-300 X 6-8 /t; ascospores 
acicular, hyaline, multiseptate, 180-200 X i'5 /*. Paraphyses branched, filiform 
with spherical heads. 

DIATRYPACEAE 

DiATRYPE CHLOROSARCA Berk. & Broomc in Journ. Linn. Soc. Lond., Bot. 14 : 123 

(1873). 

Nepal: Tumlingtar, Arun ravine, on dead twigs, 23 Dec. 1961, Norkett go6o. 

Distribution : Ceylon and India. 

The Nepal collection agrees with the type material described from Ceylon : spores 
7-9 X 2-2-5 /*. hyaline to light brown. 

HYPOCREACEAE 

cf. Balansia andropogonis Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 9 : 

395 (191 1). — Patel, Gokh. & Kulk. in Indian Phytopath. 4 : 65 (1951). 

Nepal: Mayangdi Khola, 1,000 m., on inflorescence of Chrysopogon aciculatus, 
4 Sept. 1954, Stainton, Sykes & Williams 4137. 

Distribution (of B. andropogonis) : India, Philippines. 

The inflorescence of the host is so deformed and shrouded by the fungus as to 
give the appearance of quite a different genus of grass. Only the Ephelis stage of 
the fungus is represented: conidia acicular, 20-24 X i'5 /^. 



FUNGI OF RECENT NEPAL EXPEDITIONS 105 

Epichloe cinerea Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : iii 

(1873)- 

Nepal: Taplejung district, Dhankuta Province, Sanghu, 2,000 m., on an indeter- 
minable grass haulm, 17 Oct. 1961, Norkett 56g6 C. 

Distribution: Previously recorded from Ceylon. Sydow and Butler's record from 
Mysore (Ann. Mycol. 9 : 394 (1911)) appears to be a different species, as Petch (Ann. 
Roy. Bot. Gdns. Peradeniya 7 : 88 (1920)) has noted. 

phyllachoraceae 

Catacauma repens (Corda) Theiss. & Syd. in Ann. Mycol. 13 : 383 (1915)- 

Sphaeria repens Corda, Icones Fung. 4 : 42, tab. 9 fig. 123 (1840). 
Phyllachora repens (Corda) Sacc, Syll. Fung. 2 : 597 (1883). 

Nepal: Phewa Tal, 800 m., on fallen leaves oiFicus religiosa, 6 May 1954, Stainton, 
Sykes & Williams 5254. 

Distribution: Cuba, Natal, India. 

Several species of Catacauma have been described from Ficus spp. but this Nepal 
collection is quite typical of C. repens. 

XYLARIACEAE 

Hypoxylon MULTIFORME (Frics) Fries, Summa Veg. Scand. : 384 (1849). 
Sphaeria multiformis Fries, Syst. Mycol. 2 : 334 (1823). 

Nepal: Taplejung, Mewa Khola, 2,750 m., on moss-covered branch (? birch), 
22 Jan. 1962, Norkett gjoo. Ganesh Himal, Ankhu Khola, 2,130 m., on rotting 
log, 12 May 1962, Stainton 36g4. 

Distribution: Widespread in Northern hemisphere; previously recorded from 
Nepal. 

Hypoxylon truncatum (Schwein.) J. H. Mill, in Trans. Brit. Mycol. Soc. 17 : 130 
(1932); Monogr. World Sp. Hypoxylon : 95 (1961). 

Sphaeria truncata Schwein. in Schr. Naturf. Ges. Leipz. i : 44 (1822). — Fries, Syst. Mycol. 
2 : 442 (1823). 

Nepal: Bakhri Kharka, north of Pokhara, 2,000 m., 25 Apr. 1954, Stainton, Sykes 
& Williams 5065. 

Distribution: Tropical and semitropical ; America, Africa, China, Japan. 

Miller (loc. cit., 1961) should be consulted for an account of this species and its 
considerable synonymy. 

Xylosphaera hypoxylon subsp. adscendens (Fries) Dennis in Bull. Jard. Bot. 
Bruxelles 31 : 124 (1961). 

Sphaeria adscendens Fries in Linnaea 5 : 537 (1830). 
Xylosphaera adscendens (Fries) Dennis in Kew Bull. 13 : 102 (1958). 
Xylaria hypoxylon f. tropica H. & P. Syd. & Butl. in Ann. Mycol. 9 : 418 (1911)- 
Xylaria hypoxylon var. tropica (H. & P. Syd. & Butl.) Balf.-Browne in Bull. Brit. Mus. 
(Nat. Hist.), Bot. i : 216 (1955). 
BOT. 4, 3. 7§ 



io6 FUNGI OF RECENT NEPAL EXPEDITIONS 

Nepal: South of Gurjakhani, 3,500 m., 18 Aug. 1954, Stainton, Sykes 6- Williams 

3897- 

Distribution: South America, West Indies, Africa, India, Nepal, Indonesia. 

This is the fungus previously recorded from Nepal as Xylaria hypoxylon var. 

tropica (Balfour-Browne, loc. cit.). Dennis (loc. cit., 1961) considers that it should 

be treated as a separate subspecies of Xylosphaera hypoxylon Dumort. Dennis had 

not seen the type specimen of Xylaria hypoxylon forma tropica and left open the 

question of whether that name was a synonym of Xylosphaera hypoxylon subsp, 

adscendens. I too have not seen the type, but I follow Dennis in considering that 

this is a separate subspecies. 

Xylosphaera mellissii (Berk.) Dennis in Kew Bull. 13 : 104 (1958) ; in Revista 
Biol. I : 186 (1958). 

Hypoxylon mellissii Berk, in Melliss, St. Helena : 379 (1875). 

Xylaria mellissii (Berk.) Cooke in Grevillea 11 : 85 (1883). 

Xylaria arbuscula Sacc. in Michelia i : 249 (1878). — J. H. Mill, in Bothalia 4 : 265 (1942). 

Nepal: Dhankuta Province, Chainpur district, Tumlingtar, Sabhaya Khola, 600 
m., " growing out of a niche in rock ", 8 Dec. 1961, Norkett 84^1 A. 

Distribution : Generally common in the tropics and subtropics, but not previously 
reported from the Himalayan region. 

The Nepal collection consists of branched stromata with long slender stalks and 
short cylindrical heads ; ascospores are 14-16 X 4-5 /i. 



Xylosphaera polymorpha (St. Amans) Dumort., Comment. Bot. : 92 (1822). — 
Dennis in Bull. Jard. Bot. Bruxelles 31 : 140 (1961). 

Sphaeria polymorpha St. Amans, Fl. Agenaise : 520 (1821). > 

Xylaria polymorpha (St. Amans) Grev., Fl. Edinensis : 355 (1824). 

Nepal: Arun Valley, Sabhaya Khola, 2,300 m., on tree trunk in forest, 7 Sept. 
1956, Stainton i6oy. 

Distribution : Worldwide. 

For a recent account and complete synonymy see Dennis (loc. cit.). 



Xylosphaera telfairii (Berk.) Dennis in Kew Bull. 13 : 106 (1958) ; in Bull. Jard. 
Bot. Bruxelles 31 : 119 (1961). 

Sphaeria telfairii Berk, in Ann. Nat. Hist. 3 : 397 (1839). 

Xylaria telfairii (Berk.) Fries in Nov. Acta Reg. Soc. Sci. Upsal., Ser. 3, i : 127 (1851). 

Nepal: Arun Valley, Kasuwa Khola, 2,800 m., on tree trunk in forest, 11 Sept. 
1956, Stainton i6ig. Arun Valley, Hatiar, 2,600 m., 21 Aug. 1956, Stainton 
1408. 

Distribution: Africa, Ceylon, India, Indonesia, Australia. 

For a recent account of the species, Dennis (loc. cit.) should be consulted. 



FUNGI OF RECENT NEPAL EXPEDITIONS 107 

UsTULiNA DEUSTA (Fries) Petrak in Ann. Mycol. 19 : 279 (1921). 

Sphaeria deusta Fries, Syst. Mycol. 2 : 345 (1823). 

Ustulina vulgaris Tulasne frat., Sel. Fung. Carp. 2 : 23, tab. 3 figs. 1-6, (1863), nom. 
super fl. 

Nepal: Bakhri Kharka, north of Pokhara, 2,000 m., 25 April 1954, Stainton, 
Sykes & Williams 5056. 
Distribution : Worldwide. 

coronophoraceae 
CoRONOPHORA EPISTROMA Syd. apud. Syd., Mitter & Tand. in Ann Mycol. 35 : 231 

(1937)- 

Nepal: Sanghu, guUey below camp, 2,000 m., on dead twig, 21 Dec. 1961, Norkett 
6390. 

Distribution: Originally recorded from Allahabad, and not reported since. 

This species was originally described as parasitic in the stromata of Haplosporella 
phyllanthina and again in the present gathering it is growing within the remains of 
a fungal pycnidium, but the identity of the latter could not be determined, nor that 
of the host twigs. 

PLEOSPORACEAE 

Fenestella fenestrata (Berk. & Broome) Schroet. in Cohn, Krypt.-Fl. Schles. 
3(2) : 435 (1897). (Fig. 3). 

Valsa fenestrata Berk. & Broome in Ann. Mag. Nat. Hist., Ser. 3, 3 : 366, t. 10 fig. 14 

{1859). 
Fenestella princeps Tulasne frat., Sel. Fung. Carp. 2 : 207 (1863), nom. superfl. 

t Nepal: Dhankuta district, Chitre, on dead twig, 2,000 m., 20 Sept. 1961, Norkett 
5159 A. 

Distribution: N. America, Europe. 

I The fungus occurs on an unidentified dicotyledonous twig. The pseudothecia 
are superficial, ± stipitate, arising in small groups on an inconspicuous stroma. 
Asci cyhndrical, 250 X 20 ^ approx. Ascospores 30-50 X 12-14 //, broadly fusi- 
form, dark brown, the end cells being ± hyaline ; there are 3 main and several lesser 
transverse septa and several longitudinal septa. 

Petrak (Sydowia 8 : 165 (1954)) describes Cucurbitaria pakistanica from Choa 
Saidan Shah, on Acacia modesta. This resembles the Nepal fungus in general 
structure but its spores are considerably smaller, the average size being 13-22 X 
8-10 /I. 

venturiaceae 

Rehmiodothis osbeckiae (Berk. & Broome) Theiss. & Syd. in Ann. Mycol. 12 : 192 
(1914). 

Dothidea osbeckiae Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : 134 (1873). 



I 



io8 



FUNGI OF RECENT NEPAL EXPEDITIONS 




Fig. 3. Fenestella fenestrata (Berk. & Broome) Schroet. a, asci containing 
ascospores; b, ascospore; c, stroma with pseudothecia. Norkett 51 5g A. 

Nepal: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., on Osbeckia 
sp., 4 Oct. 1961, Norkett ^6g6 D. 

Distribution: Previously recorded only from Ceylon. 

POLYSTOMELLACEAE 

ScHNEEPiA Sp. cf. DiscoiDEA (Racib.) Racib. ex Theiss. & Syd. in Ann. Mycol. 13 : 
203 (1915). 

Parmularia discoidea Racib., Parasit. Algen & Pilze Java's 2:21 (1900). 

Nepal: Dhankuta Province, Taplejung district, Sanghu, 1,600 m., on fern, 
Davallia sp., 7 Jan. 1962, Norkett 8552. 

Distribution (of 5. discoidea) : Previously recorded on Polypodium longissimum 
from Java. 



FUNGI OF RECENT NEPAL EXPEDITIONS 109 

The Nepal material is not quite ripe, but appears to represent the above species. 
Its rounded black stromata occur on the underside of the fronds and are somewhat 
raised in the centre. The asci and the i-septate brownish spores, which are not 
fully mature, measure approximately 35 X 8 /^ and 10 X 4 /* respectively. 

Several black stromatic Ascomycetes with i-septate spores have been described 
on ferns. Careful comparison of type and authentic material is needed and would 
probably reveal that a number of names are synonyms. 



BASIDIOMYCETES 

USTILAGINACEAE 

CiNTRACTiA CARicis (Pers.) Maguus in Verh. Bot. Ver. Prov. Brandenb. 37 : 79 
(1896). 

Uredo caricis Pers., Syn. Meth. Fung. : 225 (1801). 

Nepal: Lamjung Himal, 4,500 m., on inflorescence of Cyperaceae, 14 July 1954, 
Stainton, Sykes & Williams 6350. East of Chalike Pahar, 4,500 m., 4 Aug. 1954, 
Stainton, Sykes & Williams 3743. 

Distribution: Worldwide, 

Farysia olivacea (DC.) H. & P. Syd. in Ann. Mycol. 17 : 41 (1919). 
Uredo olivacea DC, Fl. Fran9. 6 : 78 (1815). 

Nepal: Bhurungdi Khola, 1,600 m., smut balls in inflorescence of Carex cruciata 
Wahl., 20 May 1954, Stainton, Sykes & Williams 5336. Near Lumsum, 2,300 m., 
24 Oct. 1954, Stainton, Sykes & Williams gi2y. 

Distribution: Worldwide. 

Sphacelotheca hydropiperis (Schumach.) de Bary, Vergl. Morph. Biol. Pilze : 
187 (1884). 

Uredo hydropiperis Schumach., Enum. PI. Saell. 2 : 234 (1803). 

Nepal: Southwest of Gurjakhani, 3,300 m., in flower of Polygonum campanulatum, 
20 Oct. 1954, Stainton, Sykes & Williams gi04. Dhankuta Province, Taplejung 
district, Sanghu, 2,000 m., 25 Nov. 1961, Norkett 7532 A. 

Distribution: Worldwide on Polygonum sp. Previously recorded from Nepal in 
1955- 

Sphacelotheca reiliana (Kuhn) Clint, in Journ. Mycol. 8 : 141 (1902). 

Ustilago reiliana Kiihn in Rabenh., Fung. Eur. : No. 1998 (1875). 

Nepal: Murigurja Gad, 2,500 m., on Zea mays, 27 July 1954, Stainton, Sykes <S 
Williams 36^3. 
Distribution: Recorded in most countries where maize is grown. 



no FUNGI OF RECENT NEPAL EXPEDITIONS 

UsTiLAGO BiSTORTARUM (DC.) Kom. in Hedwigia i6 : 38 (1877). 
Uredo bistortarum DC, Fl. Fran9. 6 : 76 (1815). 

Nepal: East of Chalike Pahar, 4,830 m., on leaves of Polygonum affine, 22 Sept. 
1954, Stainton, Sykes & Williams 4541. 

Distribution : Worldwide. 

Forming in this material elongated pustules over the under surface of the leaf 
lamina, and not occurring on the margins. The spores are pale purple, angular to 
globose, 10-16 /Ji, minutely verrucose. 

This species, common in most parts of the world, appears not to have been re- 
corded previously from India or neighbouring countries. 

Ustilago hordei (Pers.) Lagerh. in Mitt. Badischen Bot. Ver. [2] (59) : 70 (1889), 
non Ustilago hordei Bref. (1888). 

Uredo segetum var. hordei Pers., Syn. Meth. Fung. : 224 (1801). 

Nepal: Gurjakahni, 2,800 m., on barley inflorescence, i June 1954, Stainton, 
Sykes & Williams 2g4i. Tegar, north of Mustang, 6 Aug. 1954, Stainton, Sykes & 
Williams 22^2. 

Distribution: Worldwide. 

The name used for this species is illegitimate, being a later homonym of Ustilago 
hordei Bref. (in Nachr. Klub Landwirthe Berl. 221 : 1593 (28 June 1888)), which 
was based on the loose smut of barley, as a name for which it antedated U. nuda 
(Jensen) Kellerm. & Swingle (see below). Before the correct names of these two 
species can be estabhshed, it is necessary to fix the application of the name Uredo 
segetum Pers. (Syn. Meth. Fung. : 224 (1801)), and also to investigate the status 
and application of the names Uredo carho-tritici and Uredo carbo-hordei quoted by 
Kellerman & Swingle (in Annu. Rep. Exp. Stat. Kansas State Agric. Coll. 2 : 262, 
269, 278 (1890)) from a work by Philippar not available to me. I therefore adopt 
for the present the names in current use. 

Ustilago morinae Padw. & Azmat. Khan in Mycol. Pap., Imp. Mycol. Inst. 10 : i 

(1944)- 

Nepal: Above Seng Khola, 4,500 m., in flowers of Morina, 25 June 1954, Stainton, 
Sykes & Williams 32^4. 

Distribution: Recorded from Kashmir in 1944 on Morina longifolia. 

Ustilago nuda (Jens.) Kellerm. & Swingle in Annu. Rep. Exp. Stat. Kansas State 
Agric. Coll. 2 : 277 (1890). 

Ustilago segetum var. nuda Jens, in Journ. Roy. Agric. Soc, Ser. 2, 24 : 406 (1888). 
Ustilago nuda Rostr. in Tidsskr. Landokon 8 : 745 (1889), nom. nud. 

Nepal: Chimgoan, north of Tukucha, 3,000 m., on barley in field, 3 June 1954, 
Stainton, Sykes & Williams ^04. Gurjakhani, 2,830 m., on wheat, i June 1954, 



FUNGI OF RECENT NEPAL EXPEDITIONS iii 

Stainton, Sykes & Williams 2g40. Tamur valley, Mewa Khola, on wheat, 17 May 
1956, Stainton 343. 

Distribution: Worldwide. 

For the status of this name, see under Ustilago hordei above. 

UsTiLAGO piPERi Clint, in Proc. Boston Soc. Nat. Hist. 31 : 382 (1904). — G. W. 
Fisch., Man. N. Amer. Smut Fungi : 291, fig. 123 D (1953). 

Nepal: Near Seng Khola, 4,600 m., on Polygonum rumicifolium, 9 Aug. 1954, 
Stainton, Sykes & Williams 38og. 

Distribution : United States of America. 

The Nepal collection agrees well with this species; it forms large pustules on the 
underside of the leaves; the spores are pinkish purple, spherical or subspherical, 
7-9 fi diam., marked with striae. This species differs from U. bistortarum in the 
distinctly smaller spores. 

TILLETIACEAE 

Melanotaenium selaginellae Henn. & Nyman in Warb., Monsunia i : 2 (1900). 

Nepal: Dhankuta Province, Taplejung district, Sanghu, on Selaginella sp., 27 
Oct. 1 961, Norkett 6551 A. 

Distribution: Previously recorded from Tjibodas in Java. 

In the present collection the smut occurs mostly on the basal portion of the leaves, 
but occasionally it covers the whole leaf surface. Spores dark brown, globose, 
coarsely verrucose, 15-18 yM diameter. 

melampsoraceae 

CoLEOSPORiUM barclayense Bagchee in Ind. Forest Rec, New Ser., Bot. 4 : 53 

(1950). 

Nepal: Annapurna Himal, Mardi Khola, 4,160 m., on Senecio alatus Wall, ex 
DC, 20 Sept. 1954, Stainton, Sykes & Williams 8328. 

Distribution: India. 

The teleutospore stage is represented in this collection. Morphologically, the 
Coleosporium species on Senecio are difficult to separate, but, on the basis of inocula- 
tion experiments carried out by Bagchee (loc. cit.) on rusts on different species of 
Senecio, that on 5. alatus would appear to be his C. barclayense. 

Coleosporium campanulae (Pers.) Kickx, Fl. Crypt. Flandres 2 : 54 (1867). — 
Gaiim. in Beitr. Krypt.-Fl. Schweiz 12 : 113, fig. 99-100 (1959). 

Uredo campanulae Pers., Syn. Meth. Fung. : 217 (1801). 

Nepal: Near Jagat, 2,800 m., on Campanula sp., 5 July 1954, Stainton, Sykes & 
Williams 3376. Above Lumsum, 3,700 m., on Campanula seedlings, 23 Oct. 1954, 
Stainton, Sykes & Williams gi22. Rambrong, Lamjung Himal, 3,000 m., on Lobelia 
seguinii var. doniana (Skottsb.) Wimm., 27 Oct. 1954, Stainton, Sykes 6- Williams 



112 FUNGI OF RECENT NEPAL EXPEDITIONS 

8306. Annapurna Himal, Mardi Khola, 2,600 m., on Lobelia sp., 20 Sept. 1954, 
Stainton, Sykes & Williams 8531. Baglung, 1000 m., on Wahlenbergia gracilis DC, 
20 Apr. 1954, Stainton, Sykes & Williams 548. 

Distribution: Widely represented in the Northern hemisphere and previously 
recorded from Nepal. 

The uredospores occurred on stems as well on both leaf surfaces. A greater 
proportion of the spores on Wahlenbergia were oblong and elongated compared with 
those on Campanula. Probably two biological forms are represented. Gaiimann 
(loc. cit.) should be consulted for an account of at least six such forms which have 
been isolated from different genera and species of Campanulaceae. 

CoLEOSPORiUM INULAE Rabcnh. in Bot. Zeit. 9 : 455 (1851). — P. & H. Syd., Monogr. 
Ured. 3 : 609 (1915). — Vasud. in Butl. & Bisby, Fungi of India, Rev. Ed. : 85 
(i960). 

Uredo inulae Kunze in Klotsch, Herb. Viv. Mycol., No. 589 (1844), nom. nud. 

Nepal: Pasgam, 1,500 m., on Inula cappa DC, 25 June 1954, Stainton, Sykes & 
Williams 5938. 

Distribution: Europe, North Africa, Asia Minor, India. 

The uredo stage is represented and this chiefly on the upper surface of the leaves, 
with very few pustules on the lower. 

COLEOSPORIUM PEDicuLARiDis Tai in Farlowia 3 : 100 (1947). 

Nepal: Taglung, south of Tukucha, Kali Gandaki, 4,000 m., on leaves of Pedicu- 
laris sp., 22 Sept. 1954, Stainton, Sykes & Williams yg8y. Panchasi, 2,300 m., on 
Pedicularis sp., 15 Oct. 1954, Stainton, Sykes & Williams 8944. 

Distribution: China. 

There is httle doubt that the Nepal fungus represents this species. The uredosori 
are chiefly on the under surface of the leaves, light yellow and soon pulverulent. 
The uredospores are coarsely and densely echinulate, 18-30 X 12-20 /*, oblong to 
irregularly globose. 

CoLEOSPORiUM PLECTRANTHi Barcl. in Journ. Asiatic Soc. Bengal 59 (2): 89 (1890). 

Nepal: Sattewati, 2,000 m., on Plectranthus sp., 12 Oct. 1954, Stainton, Sykes & 
Williams 8950. 

Distribution: India, Japan. 

Coleosporium senecionis (Pers.) Fries, Summa Veg. Scand. : 512 (1849). 
Uredo farinosa var. senecionis Pers., Syn. Math. Fung. : 218 (1801). 

Nepal: Above Lumsum, 3,160 m., on Senecio graciliflorus DC, 10 Sept. 1954, 
Stainton, Sykes & Williams 432^. Dhankuta Province, Milke Danda, 3,160 m., 
on Senecio graciliflorus DC, 16 Nov. 1961, Norkett yi2i A. 

Distribution : Worldwide. 

Uredospore stage. See note under Coleosporium barclayense. 



FUNGI OF RECENT NEPAL EXPEDITIONS 113 

Hyalopsora polypodii (Pers.) Magn. in Ber. Deutsch. Bot. Ges. 19 : 582 (1902). 

Uredo linearis var. polypodii Pers., Syn. Meth. Fung. : 217 (1801). 

Nepal: Annapurna Himal, Seti Khola, 4,160 m., on Polypodium malacodon, 
3 Aug. 1954, Stainton, Sykes 6- Williams 6603. 

Distribution: North America, Europe, India, Japan. 

Melampsora sp. cf. HiRCULi Liudr. in Acta. Soc. Fauna & Flora Fenn. 22 (3) : 19 

(1902). 

Nepal: Annapurna Himal, 4,150 m., on Saxifraga moorcroftiana Wall, 2 Aug. 
1954, Stainton, Sykes & Williams 6^8y. 

Distribution (of M. hirculi) : Europe (Finland, Russia, Switzerland) . 

Although this species so far has apparently only been recorded from Europe, the 
Nepal collection agrees very closely with the original description of M. hirculi, but 
the fungus occurs not only on the under surfaces, but more frequently on the upper 
leaf surfaces, which become considerably discoloured and blotchy. Uredospores are 
globose, ovate or elhpsoid, minutely verrucose, 16-25 X 14-18 /^; paraphyses are 
abundant and capitate or clavate, 35-60 X 10-20 jn with a thick wall. The teleuto- 
spores are brown, oblong, 35-55 X 8-12 jn. 

Milesina sp. cf. EXIGUA FauU in Journ. Arnold Arb. 12 : 218 (1931). 

Nepal: Dhankuta Province, Taplejung district, 1,950 m., on Diplazium sp., 
9 Jan. 1962, Norkett 8642 A. 

Distribution (of M. exigua) : Poland, Japan and the Siberian coast. 

Uredosori only are represented; they occur in brown discoloured areas on both 
sides of the fronds. Many of the spores are somewhat irregularly polygonal in 
shape. They measure 28 X 15 /^ (average) and are quite smooth. 

PUCCINIACEAE 

Frommea duchesneae (Arth.) Arth. in Bull. Torrey Bot. Club 44 : 504 (1917) ; in 
N. Amer. Fl. 7 : 731 (1925). 

Kuhneola duchesneae Arth. in N. Amer. Fl. 7 : 185 (1912). 

Frommea obtusa [var.] duchesneae (Arth.) Arth., Man. Rusts U.S. & Canada : 93 (1934). — 

Gaum, in Beitr. Krypt.-Fl. Schweiz 12 : 1177 (i959)- 
Frommea obtusa-duchesneae Vienn.-Bourg. in Rev. Path. V6g. Entom. Agric. Fr. 33 : 38 

(1954), worn, superfl. 

Nepal: Ghar Khola, 1,600 m., on Duchesnea indica, 14 June 1954, Stainton, Sykes 
& Williams Sy6j. 

Distribution: Nepal, and North America and France, where the host is naturahzed. 

It seems that Arthur in 1934 (loc. cit.) had second thoughts about treating the 
rust on Duchesnea indica as a full species and, describing it as a " less robust form " 
compared with F. ohtusa, regarded it merely as a variety of the latter species. 

In the Nepal collection the uredo-sori are abundant, the uredospores measure 

BOX. 4, 3. 7§§ 



114 FUNGI OF RECENT NEPAL EXPEDITIONS 

9-14 X 18-20 fi, distinctly small for F. obtusa (which occurs on Potentilla sp.) but 
in agreement with the spores described on Duchesnea from America by Arthur and 
from France by Viennot-Bourgin. No other spore form was present on the Nepal 
plants. In the absence of firm evidence to support synonymy with F. obtusa the 
name F. duchesneae is retained here. This is the first record of the rust in what is 
considered to be the host plant's centre of origin. 

Gymnosporangium padmarense Balf.-Browne in Bull. Brit. Mus. (Nat. Hist.) i : 

205, fig. 2 (1955)- 

Nepal: Near Gurjakhani, 3,000 m., on Juniperus wallichiana, 3 June 1954, 
Stainton, Sykes & Williams 2g6g. 

Distribution: Nepal. 

The material under consideration is old, the sori are broken up and the teleuto- 
spores have lost their pedicels. 

Gymnosporangium cunninghamianum Barcl. in Sci. Mem. Med. Off. Army Ind. 

5 : 78, t. 1-3 (1890). 

Nepal: Village south of Chakure Lekh, 6 Apr. 1952, Polunin, Sykes & Williams 
i8yo. 

Distribution: India, Nepal. 

Phragmidium incompletum Barcl. in Journ. Asiatic Soc. Bengal 59 (2) : 83 (1890). 

Nepal: Siklis, north of Pokhara, 3,000 m., on Rubus sp., 21 Apr. 1954, Stainton, 
Sykes & Williams 4950. 

Distribution: India. 

The uredospore stage only is present, and it agrees well with the original account, 
except that it is chiefly epiphyllous. Uredospores have a thick, 3-4 /i, epispore, 
which is warted. There are no paraphyses. 

Phragmidium sp. cf. nepalense Barcl. in Journ. Asiatic Soc. Bengal 60 (2) : 220 

(1891). 

Nepal: Bhuji Khola, 2,800 m., on Potentilla nepalensis, 16 Oct. 1954, Stainton, 
Sykes & Williams go^S. 

Distribution (of P. nepalense) : India. 

Uredosori only are represented. In the original description it was not stated 
whether the epispore is smooth or echinulate. Padwick & Azmatullah Khan 
(Mycol. Papers, Imp. Mycol. Inst. 10 : 4 (1944)) record this species and describe the 
uredospores as finely echinulate. This agrees with those on the present collection, 
and the measurements are similar. 

Puccinia caricis var. himalayensis (Barcl.) Padw. & Azmat. Khan in Mycol. 

Papers, Imp. Mycol. Inst. 10 : 9 (1944). 

Aecidium urticae var. himalayense Barcl. in Sci. Mem. Med. Off. Army Ind. 2 : 29, t. 4, 
5 figs. 8-17 (1887); in Journ. Asiatic Soc. Bengal 56 (2) : 368 (1887). 



J 



FUNGI OF RECENT NEPAL EXPEDITIONS 115 

Nepal: Jagat, 2,600 m., on Urtica sp., 5 July 1954, Stainton, Sykes & Williams 

3364- 

Distribution: India, Himalayas. 

Cf. Padwick & Azmatullah Khan (loc. cit.) for an account of the somewhat con- 
fused synonymy. 

PucciNiA FAGOPYRi Barcl. in Journ. of Bot. 28 : 261 (1890). 

Nepal: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., on Fagopy rum, 
II Nov. 1961, Norkett yiys A. 
Distribution: India. 

PucciNiA gentianae (Strauss) Link in L., Sp. PL, Ed. 4, 6 (2) : 73 (1825). 
Uredo gentianae Strauss in Ann. Wetter. Ges. 2 : 102, t. 11 fig. 33 (1811). 

Nepal: South of Gurjakhani, 4,000 m., on Gentiana sp., 8 June 1954, Stainton, 
Sykes & Williams 3oyi. 

Distribution : Widely distributed in the northern Hemisphere. 
Only the aecidial stage is represented in the Nepal material. 

PucciNiA GLUMARUM (J. K. Schmidt) Erikss. & Henn, in Medd. K. Landtbr.-Akad. 
Exp. 38 : 141 (1896). 

Uredo glumarum J. K. Schmidt in AUgem. Okonom.-Tech. Flora i : 27 (1827). 

Nepal: Gurjakhani, 2,800 m., on Triticum vulgare, 1 June 1954, Stainton, Sykes 
& Williams 2gjg. 

Distribution: Worldwide. 

PucciNiA GRAMiNis Pers., Syn. Meth. Fung. : 228 (1801). 

Nepal: Ghasa, Kali Gandaki Valley, 2,500 m., on Berberis sp., 31 May 1954, 
Stainton, Sykes & Williams 5499; same locality and host, 13 June 1954, Stainton, 
Sykes & Williams syso. 

Distribution : Worldwide. 

Some of the pustules were a bright pink and considerably swollen. 

PucciNiA LANTANAE Farl. in Proc. Amer. Acad. Arts & Sci. 18 : 83 (1883). — G. 
Laund. in Mycol. Papers, Commonw. Mycol. Inst. 89 : 43 (1963). 

Nepal: Sanghu, Dhankuta Province, 2,000 m., on Justicia diffusa, 15 Oct. 1961, 
Norkett 6124. 

Distribution: North and South America, India, Indonesia, Philippines, China, 
Japan. 

PucciNiA leucophaea H. & P. Syd. & Butl. in Ann. Mycol 10 : 258 (1912). 

Nepal: Ghar Khola, 2,000 m., on Colquhounia coccinea, 14 June 1954, Stainton, 
Sykes & Williams SysS. 



ii6 FUNGI OF RECENT NEPAL EXPEDITIONS 

Distribution: India. 

Aecidial stage only and chiefly epiphyllous ; only a few isolated sori on the lower 
surface and on the petioles. In the original account the aecia were said to be 
hypophyllous. 

PucciNiA poLYGONi-AvicuLARiAE Pers., Syn. Meth. Fung. : 227 (1801). 

Puccinia polygoni Alb. & Schwein., Consp. Fung. : 132 (1805), nom. superfl. — Gaum, in 
Beitr. Krypt.-Fl. Schweiz 12 : 775 (1959). 

Nepal: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., on Polygonum 
nepalense, 25 Nov. 1961, Norkett 7530 A. 

Distribution : Worldwide. 

The teleutospore stage is represented. This species and P. polygoni-amphibii are 
united by some mycologists but treated as separate by others. The two species 
are maintained here not only on biological grounds but also on account of definite 
small morphological distinctions, which apply not only to European plants, as 
generally stated, but also to American specimens as exemplified in the British 
Museum herbarium. In P. polygoni-amphibii, usually occurring on Polygonum 
amphihium, P. lapathifolium and related species, the teleutosori remain for long 
covered by the epidermis and form small pimply pustules, and the spores are 
frequently somewhat bent and easily lose their pedicels. On the other hand, in P. 
polygoni-aviculariae, usually on Polygonum dumetorum and P. convolvulus, the 
pustules rapidly burst through the epidermis of the host, leaving smooth black 
cushions of straight, stalked teleutospores. 

In the present instance the host, Polygonum nepalense, resembles in general 
appearance and texture Polygonum convolvulus, and the teleutospores and sori of 
the fungus agree exactly with those described for P. polygoni-aviculariae. 

Puccinia pulverulenta Grev., Fl. Edin. : 432 (1824). — Gaiim. in Beitr. Krypt.-Fl. 

Schweiz 12 : 929 (1959). 

Nepal: Near Dogadi Khola, 4,300 m., on Epilohium sp., 23 June 1954, Stainton, 
Sykes & Williams 3226. 

Distribution: Worldwide. 

Puccinia ustalis Berk, in Hook., Journ. Bot. 6 : 207 (1854). i 

Puccinia songarica Jacz. in Hedwigia 39 : (130), fig. i (1900). 

Nepal: Rambrong, Lamjung Himal, 4,000 m., on Ranunculus sp., 29 June 1954, 
Stainton, Sykes & Williams 6016. 

Distribution: India, Turkestan, Mongolia. 

Ravenelia emblicae Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 4 : 438 (1906). 

— P. & H. Syd., Monogr. Ured. 3 : 293 (1914). 

Nepal: Chainpur path, Tumlingtar, 800 m., on Phyllanthus emblica, 21 Dec. 1961, 
Norkett 8100. 



FUNGI OF RECENT NEPAL EXPEDITIONS 117 

Distribution: Previously recorded from India and Burma. 

Ravenelia phyllanthi Mundk. & Thirum. (Mycol. Papers, Imp. Mycol. Inst. 16 : 
24, iig. 18 (1946)) seems to be synonymous. This was described on Phyllanthus 
polyphyllus from Mysore. 



UREDIN ALES— Form. Genera 

Aecidium crini Kalchbr. in Grevillea ii : 26 (1882). — -Mundk. & Thirum. in Mycoh 
Papers, Imp. Mycol. Inst. 16 : 16 (1946). 

Aecidium amaryllidis H. & P. Syd. & Butl. in Ann. Mycol. 10 : 274 (1912). 

Nepal: Dana, Kali Gandaki Valley, 1,600 m., on Crinum amoenum Roxb. ex 
Ker-Gawl., 13 June 1954, Stainton, Sykes & Williams 5738. 
Distribution : India, South Africa. 

Aecidium infrequens Barcl. in Journ. Asiatic Soc. Bengal, 59 (2) : 105 (1890). 

Nepal: Nr. Dogadi Khola, 4,300 m., on Geranium collinum Steph. ex Willd., on 
open slopes, 23 June 1954, Stainton, Sykes & Williams 3227. 

Distribution: India, Nepal, Japan. 

The aecidia cover considerable areas of the lower side of the leaves, showing as 
light brown patches on the upper surface. Each aecidium is 250-300 ju, diam. 
The size, ornamentation, colouring of the aecidia and aecidiospores correspond 
exactly with those originally described for A. infrequens by Barclay on a Geranium 
sp. (? nepalensis) from Simla. The present fungus agrees also with the details 
given for A . sanguinolentum on other Geranium spp. in Finland, Russia and America by 
Lindroth (Bot. Notiser 1900 : 241), and he suggested this might be a synonym of 
Barclay's rust. 

Polunin, Sykes & Williams 4765, recorded in 1955 in this Journal (Balfour-Browne, 
1955) as A. infrequens is microscopically similar to the fungus now reported, but 
differs in that the aecidia are grouped in orbicular patches with a small bare spot 
in the centre of each patch, i.e., as described by Lindroth for the less heavily in- 
fected specimens of his A . sanguinolentum. The host of this earlier collection, which 
has now been identified also as G. collinum, is more elegant and slender. Possibly 
the heavy rust infection is responsible for the coarser growth of the host specimen 
of Stainton, Sykes & Williams 3227. 

A. sanguinolentum, in consequence of inoculation experiments from Geranium spp. 
of European origin (Lindroth, loc. cit.), has been described as a stage in life history of 
Puccinia polygoni-amphihii Pers. However, until experiments are made using 
specimens of the Nepal fungus, no certain conclusions can be drawn as to its relation- 
ships or alternative host plants. 

In the meantime A. infrequens Barcl. is the name preferred on grounds of distri- 
bution and would be also on grounds of priority should this species prove to be 
identical with A . sanguinolentum. 



Ii8 FUNGI OF RECENT NEPAL EXPEDITIONS 

Aecidium SCUTELLARIAE Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 5 : 504 
(1907). 

Nepal: Ghar Khola, 1,800 m., on Scutellaria scandens D. Don., 14 June 1954, 
Stainton, Sykes & Williams 3763. 

Distribution: Himalayas. 

A similar fungus, Aecidium scutellariae-indicae Dietel, has been described from 
Japan on Scutellaria indica var. japonica. This may be identical with the above 
species (P. & H. Syd., Monogr. Ured. 4 : 115 (1923)). 

Peridermium orientale Cooke in Ind. Forester 3 : 91 (1877) " orientalis ". 
Aecidium complanatum Barcl. in Journ. Asiatic Soc. Bengal 59 (2) : loi (1890). 

Nepal: Dhaibungkot, 1,600 m., on dead pine needles, 31 May 1949, Polunin 041. 
Near Beni, 1,300 m., on needles of Pinus longifolia, 23 May 1954, Stainton, Sykes & 
Williams 2yg4. 

Distribution: India, Nepal, Bhutan. 

Uredo hyperici-mysorensis Petch in Ann. R. Bot. Gdns. Peradeniya 6 : 213 

(1917)- 

Nepal: Sanghu, Dhankuta Province, 1,400 m., on Hypericum sp., 12 Nov. 1961, 
Norkett yo88. 

Distribution: Ceylon, India. 

auriculariaceae 

AuRicuLARiA DELiCATA (Fries) Hcnn. apud Bresad., Henn. & Magn. in Engl., Bot. 
Jahrb. 17 : 492 (1893). 

Laschia delicata Fries in Linnaea 5 : 533 (1830). 

Laschia tremellosa Fries, Summa Veg. Scand. : 325 (1849). 

Nepal: Ranipauwa, north of Beni, Kali Gandaki, 1,000 m., 3 Sept. 1954, Stainton, 
Sykes & Williams y62g. 

Distribution: mostly tropical; America, Africa, India, Australia, Pacific. 

AuRicuLARiA mesenterica Pcrs., Mycol. Eur. i : 97 (1822). 

Nepal: Arun Valley, Num, 1,500 m., on tree trunk, 30 Aug. 1956, Stainton 1459. 

Distribution: America, Europe, India, Indonesia, Australia. 

The fructifications are broadly attached, many more or less disciform, and they 
therefore superficially resemble A. peltata Lloyd. However the hairs are much 
longer, up to 500 /i. 

AuRicuLARiA POLYTRiCHA (Mout.) Sacc. in Atti R. 1st Veneto, Ser. 6, 3 : 722 (1885). 

Exidia polytricha Mont, in B61ang., Voy. aux Indes-Or. 2 : 154 (1834). 
Hirneola polytricha (Mont.) Fries in K. Vet.-Akad. Handl. 1848 (i) : 146 (1849). 

Nepal: Tamrang Khola, 2,300 m., on branch of tree, 21 Nov. 1961, Norkett y88g. 
Distribution: Worldwide, 



I 



I 



FUNGI OF RECENT NEPAL EXPEDITIONS 119 

TREMELLACEAE 

GuEPiNiA HELVELLOIDES (Fries) Fries, Elenchus Fung. 2 : 31 (1828). 

Tremella helvelloides Fries, Syst. Mycol. 2 : 211 (1822). 

Phlogiotis helvelloides (Fries) Martin in Amer. Journ. Bot. 23 : 628 (1936). 

Tremella rufa Pers., Mycol Eur. i (i) : 103 (1822). 

Gyrocephalus rufus (Pers.) Bref. in Unters. Gesammtgeb. Mykol. 7 : 131 (1888). 

Nepal : Taglung, Kali Gandaki, 3,500 m., 22 Sept. 1954, Stainton, Sykes & Williams 

7990. 
Distribution : North America, Europe, China, India. 

Tremella mesenterica Fries, Syst. Mycol. 2 : 214 (1822). 

Nepal: Dhankuta Province, near Mahe, 1,300 m., on dead tree, 20 Sept. 1961, 
Norkett 5iy5 D. 

Distribution : Worldwide. 

exobasidiaceae 

EXOBASIDIUM sp. 

Nepal: Lete, Kali Gandaki Valley, 3,800 m., on Rhododendron campanulatum, 
4 June 1954, Stainton, Sykes & Williams 56oy. Above Sauwala Khola, 3,800 m., 
on Rhododendron lepidotum, 15 Sept. 1954, Stainton, Sykes & Williams 4430. Near 
Lumsum, 2,300 m., on Rhododendron seedlings, 24 Oct. 1954, Stainton, Sykes & 
Williams 9132. 

Specific identification could not be made as the collections in all cases were very 
over-ripe. Several species of Exobasidium have been described on Rhododendron. 
References to the literature on Exobasidium can be found in Sundstrom (Phytopath. 
Zeitschr. 40 : 213-17 (i960)) and in McNabb (Trans. R. Soc. N.Z., Bot. i : 267 
(1962)). 

agaricaceae 
Armillaria mellea (Fries) Kummer, Fiihr. Pilzk. : 134 (1871). 

Agaricus melleus Fries, Syst. Mycol i : 30 (182 1). 

Nepal: Arun Valley, Kasuwa Khola, on tree trunk in forest, 11 Sept. 1956, 
Stainton 1618. 

Distribution : Worldwide. 

Clitocybe tabescens (Fries) Bresad., Fung. Trident. 2 : 84, t. 197 (1900). 

Agaricus tabescens Fries, Hymenomyc. Eur., Ed. 2:111 (1874). 

Nepal: Lete, Kali Gandaki Valley, 2,600 m., in leaf mould at base oiPinus chylla, 
3 June 1954, Stainton, Sykes & Williams S55^- 
Distribution : Worldwide. 



I20 FUNGI OF RECENT NEPAL EXPEDITIONS 

COPRINUS COMATUS (Fries) Gray, Nat. Arrang. Brit. PI. i : 633 (1821). 
Agaricus comatus Fries, Syst. Mycol. i : 307 (1821). 

Nepal: Chimgaon (north of Tukucha), Kali Gandaki, 4,500 m., 17 July 1954, 
Stainton, Sykes & Williams 1846. 
Distribution : Worldwide. 

CoPRiNus DissEMiNATUs (Fries) Gray, Nat. Arrang. Brit. PI. i : 634 (1821). 

Agaricus disseminatus Fries, Syst. Mycol. i : 305 (1821). 

Psathyrella disseminata (Fries) Qu61. in M6m. Soc. fimul. Montb^liard, Ser. 2, 5 : 153 
(1872) (reimpr. quam Champ. Jura Vosg. : 123 (1872)). 

Nepal: Arun Valley, Hatiar, 2,300 m., on fallen tree in forest, 20 Aug. 1956, 
Stainton 1394. 

Distribution : Worldwide. 

Crepidotus mollis (Fries) Staude in Festg. Mitgl. XIX Versamml. deutsch. Land- 
und Forstwirthe Coburg : 71 (1857) (reimpr. quam Schwamme Mitteldeutschl. : 
71 (1858)). 

Agaricus mollis Fries, Syst. Mycol. i : 274 (1821). 

Nepal: Arun Valley, Kasuwa Khola, 2,800 m., on tree trunk in forest, 11 Sept. 
1956, Stainton i6iy. 

Distribution: America, Europe, China, Japan, Australia. 

Very badly crushed in pressing but the layers of parallel and gelatinous hyphae 
were readily observed; spores 9x5/^, smooth. 

GoMPHUS FLOCCOSUS (Schwein.) Sing, in Lloydia 8 : 140 (1945). 

Cantharellus floccosus Schwein. in Trans. Amer. Phil. Soc, New Ser. 4 : 153 (1832). 

Nepal: Above Sauwala Khola, 3,300 m., on earth bank in Quercus forest, 13 
Sept. 1954, Stainton, Sykes & Williams 4375. 

Distribution: Recorded from North America, China, Japan, as well as from Nepal 
(1955). 

Laccaria laccata (Fries) Cooke in Grevillea 12 : 70 (1884). 

Agaricus laccatus Fries, Syst. Mycol. i : 106 (1821). 

Nepal: Arun Valley, Barun Khola, 4,000 m., in short grass, pinkish brown, 
15 Sept. 1956, Stainton 1662. 
Distribution : Worldwide. 

Lactarius pubescens (Krombh.) Fries, Epicrisis Syst. Mycol. : 335 (1838). 

Agaricus pubescens Krombh., Naturg. Abbild. & Beschreib. Essb., Schadl. & Verdacht. 
Schwamme 2 : 24, t. 13 figs. 1-14 (1832). 



FUNGI OF RECENT NEPAL EXPEDITIONS 121 

Nepal: Arun Valley, Barun Khola, 4,000 m., in short grass, 15 Sept. 1956, Stainton 
1660. 

Distribution : Apparently worldwide, but it is uncertain how many records under 
the name of the coarser L. torminosus (Fries) Gray, from which many mycologists 
have not separated this species, refer to it. Very few species of Lactarius have as 
yet been recorded from India or any of the neighbouring countries. 

Lepiota erminea (Fries) Gill., Hymenomycetes : 59 (1874). 
Agaricus ermineus Fries, Syst. Mycol. i : 22 (1821). 

Nepal: Mathand, near Pokhara, 1,120 m., on shady bank, " white except top of 
cap which is brown ", 22 June 1954, Stainton, Sykes & Williams 5852. 
Distribution: Europe, India, Australia. 

Marasmius crinis-equi F. von Muell. ex Kalchbr. in Grevillea 8 : 153 (1880). 

Marasmius equicrinis F. von Muell. ex Berk, in Journ. Linn. Soc. Lond., Bot. 18 : 383 
(1881), nom. superfl. 

Nepal: Murigurja Gad, 2,500 m., on dead vegetation near ravine track, 27 July 
1954, Stainton, Sykes & Williams 364"/. 

Distribution: America, India, Ceylon, Philippines, Australia, 

Partus polychrous (Lev.) Singer ex Balfour-Browne, comb. nov. 

Lentinus polychrous L6v. in Ann. Sci. Nat., Ser. 3, Bot. 2 : 175 (1844). 

Lentinus vellereus Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 301 (1868). 

Lentinus kurzianus Currey in Trans. Linn. Soc. Lend., Ser. 2, Bot. i : 120, t. 20 fig. 11 

(1876). 
Panus polychrous Sing., Agaricales in Modern Taxonomy, Ed. 2 : 172 (1962), nom. invalid. 

Nepal: Tumlingtar, Sabhaya Khola, 600 m., on dead tree, 20 Dec. 1961, Norkett 
8108. 

Distribution: Cuba, India, Nepal, Ceylon, Philippines, Australia. 
A few additional synonyms are given by Singer (loc. cit.). 

Panus tigrinus (Fries) Sing, in Lilloa 22 : 275 (1951). 

Agaricus tigrinus Fries, Syst. Mycol. i : 176 (1821). 

Lentinus tigrinus (Fries) Fries, Epicrisis Syst. Mycol. : 389 (1838). 

Nepal: Midam Khola, Chisankhu, 650 m., on dead tree trunk, 4 May 1954, 
Stainton, Sykes & Williams 5214. 
Distribution : Worldwide. 

Pholiota squarrosa (Fries) Kummer, Fiihr. Pilzk. : 84 (1871). 

Agaricus squarrosus Fries, Syst. Mycol. i : 243 (1821). 

Nepal: Chimgaon (north of Tukucha), Kali Gandaki, 3,500 m., in forest, at base 
of conifer, 14 Sept. 1954, Stainton, Sykes & Williams 7830. 
Distribution : North America, Europe, Japan. 



122 FUNGI OF RECENT NEPAL EXPEDITIONS 

ScHizoPHYLLUM COMMUNE Fries, Syst. Mycol i : 330 (1821). 

Nepal: Ranipauwa (north of Beni), Kali Gandaki, 1,000 m., on tree, 12 Sept. 
1954, Stainton, Sykes 6- Williams y8i8. Tumlingtar, Sabhaya River, Chainpur 
district, 600 m., 12 Dec. 1961, Norkett 8480. Hinwan Khola, Chainpur, 660 m., 
on dead stick, 21 Dec. 1961, Norkett go24. 

Distribution: Worldwide. 

HYDNACEAE 

Hericium erinaceus (Fries) Pers., Mycol. Eur. 2 : 153 (1825). 
Hydnum erinaceus Fries, Syst. Mycol. i : 407 (1821). 

Nepal: Above Sauwala Khola, 3,300 m., on Quercus in thick forest, 15 Sept. 
1954, Stainton, Sykes & Williams 4415. Chimgaon, Kali Gandaki, 3,300 m., 14 
Sept. 1954, Stainton, Sykes & Williams y82i. 

Distribution: America, Europe, India, Japan, China. 

Hydnellum zonatum forma vespertilio (Berk.) Coker & Beers, Stipitate Hydnums 
of Eastern U.S. : 80 (1951). 

Hydum vespertilio Berk, in Hook., Journ. Bot. 6 : 167 (1854). 

Nepal: Taglung, Kali Gandaki, 3,500 m., 22 Sept. 1954, Stainton, Sykes & Williams 
7991. 

Distribution: America, Europe, India. 

polyporaceae 

Amauroderma rugosum (Blume & Nees) Torrend in Broteria, Ser. Bot. 18 : 127 
(1920). 

Polyporus rugosus Blume & Nees in Nov. Act. Phys.-Med. Acad. Caes. Leop.-Car. 13 : 21, 
t. 7 (1826). 

Nepal: Dhankuta Province, below Sanghu, on roots of bamboo, 1,800 m., 27 
Feb. 1962, Norkett 10233. 

Distribution: Mostly tropical; Madagascar, India, Ceylon, Java, Philippines. 

Ganoderma applanatum (Pers.) Patouill. in Bull. Soc. Mycol Fr. 5 : 67 (1889). — 
Humphr. & Lewis in Philipp. Journ. Sci. 45 : 514 (1931). 

Polyporus fomentarius var. applanatus Pers., Mycol. Eur. 2 : 80 (1825). 
Polyporus applanatus (Pers.) Wallr., Fl. Crypt. Germ. 2 : 591 (1833). 

Nepal: Bakhri Kharka, north of Bokhara, 1,800 m., on rotten tree trunk, 24 
April 1954, Stainton, Sykes & Williams 5060. Taglung, Kali Gandaki, 3,300 m., 
19 Oct. 1954, Stainton, Sykes & Williams 8198. Dharan Bazar, Terai forest, south 
of Gopa Gurkha Camp, 250 m., on old tree, 27 Feb. 1962, Norkett 10234. 

Distribution : Worldwide. 



FUNGI OF RECENT NEPAL EXPEDITIONS 123 

Ganoderma lucidum (Fries) Karst. in Rev. Mycol. 3 (9) : 17 (1881). 

Polyporus lucidus Fries, Syst. Mycol. i : 353 (1821). 

Nepal: Bakhri Kharka, north of Pokhara, 2,000 m., on rotten tree trunk, 25 
April 1954, Stainton, Sykes & Williams Soy^. 
Distribution : Worldwide 

FoMES PECTiNATUS (Klotzsch) Gill., H3anenomycetes : 686 (1874). 

Polyporus pectinatus Klotzsch in Linnaea 8 : 485 (1833). 

Nepal: Between Bakhri Kharka and Rambrong, 2,300 m., on rotten tree trunk, 
26 Apr. 1954, Stainton, Sykes & Williams 5081. 
Distribution: America, Europe, India, Australia, Philippines. 

FoMES MARGiNATUS (Fries) Gill., Hymenomycetes : 683 (1874). 

Polyporus marginatus Fries, Syst. Mycol. i : 372 (1821). 

Nepal: Taglung, Kali Gandaki, 3,300 m., on forest tree, 11 July 1954, Stainton, 
Sykes & Williams 1751. Also at 3,500 m., 22 Sept. 1954, Stainton, Sykes & Williams 

7992. 

Distribution: America, Europe, India, Nepal, China, Japan. 

Polyporus arcularius Fries, Syst. Mycol. i : 342 (1821). 

Polyporellus arcularius (Fries) Pilat in Kav. & Pilat, Atlas Champ. 3 : 75, t. 30-31, fig. 
18 (1936). 

var. arcularius. 

Nepal: Midam Khola, Chisankhu, 660 m., 4 May 1954, Stainton, Sykes & Williams 
5210. Kabre, Kali Gandaki, 2,000 m., 13 June 1954, Stainton, Sykes & Williams 
5742. Dhankuta Province, Chainpur district, Tumlingtar, 600 m., on dead trunk, 
13 Dec. 1961, Norkett 8iog B; 8815 A ; and on 14 Dec. 1961, Norkett 8ioy B. 

Distribution : Worldwide ; previously recorded from Nepal. 

var. STRiGosus Bourd. & Galz., Hymenomyc. Fr. : 532 (1928). 

Nepal: ChipU, North of Pokhara, 2,600 m., on rotten tree trunk, 18 Apr. 1954, 
Stainton, Sykes & Williams 4882. 
Distribution : Worldwide. 
Distinguished by its marginal hairs. 

Polyporus pargamenus Fries, Epicrisis Syst. Mycol. : 480 (1838). — Overh., Polyp. 

U.S., Alaska & Canada : 336 (1953). 

Nepal: Sanghu, Milke Danda Forest, 2,900 m., on old dead tree, 16 Nov. 1961, 
Norkett 7129. 



124 FUNGI OF RECENT NEPAL EXPEDITIONS 

Distribution : Widespread in temperate and tropical regions. 

The material is in good condition but not sporing. There is much confusion over 
the use of this name, P. hiformis Klotsch and P. cervinus Fries. It is hoped to make 
a more critical study of the problem shortly. Meanwhile Overholt's interpretation 
of P. pargamenus is adopted. 

PoLYPORUs CORRUGATUS Pcrs. apud Gaud, in Freyc, Voy. aut. Monde Uranie & 
Physicienne, Bot. : 172 (1826). \ 

Earliella corrugata (Pers.) Murrill in Bull. Torrey Bot. CI. 34 : 468 (1907). 
Polystictus persoonii Cooke in Grevillea 14 : 85 (1886). 
Daedalea sanguinea Klotsch in Linnaea 8 : 481 (1833). 

Nepal: Chainpur district, Tumlingtar, Dhankuta Province, 600 m., on dead tree, 
13 Dec. 1961, Norkett 8iog E. 

Distribution: West Indies, India, Nepal, East Indies and throughout most of the 
tropics. 

PoLYPORUS PiciPES Fries, Epicrisis Syst. Mycol. : 440 (1838). — Overh., Polyp. U.S., 
Alaska & Canada : 262 (1953). 

Polyporellus picipes (Fries) Karst. in Bidr. Kann. Finl. Natur. & Folk 37 : 31 (1882). — 
Pilat in Kav. & Pildt, Atlas Champ. 3 : 99, t. 44 fig. 1-3, t. 46 fig. b, fig. 24, p. 105 
fig. B (1937)- 

Nepal: Ghar Khola, 3,100 m., on dead trunk, 15 June 1954, Stainton, Sykcs & 
Williams 5y6g. Arun Valley, Kasuwa Khola, 2,800 m., 11 Sept. 1956, Stainton 
1624. 

Distribution: Worldwide. 

PoLYPORUS SQUAMosus Fries, Syst. Mycol. i : 343 (1821). 

Nepal: Rambrong ridge, north of Pokhara, 3,300 m., on rotten tree trunk, 27 
Apr. 1954, Stainton, Sykes & Williams 5103. 
Distribution : Worldwide. 

PoLYPORUS sulphureus Fries, Syst. Mycol. i : 357 (1821). 

Grifola sulphurea (Fries) Pildt in Beih. Bot. Centralbl. 52 (B) : 39 (1934). 
Laetiporus sulphureus (Fries) Bondartz. & Sing, in Ann. Mycol. 39 : 51 (1941). 

Nepal: Chimgaon, KaU Gandaki, 3,500 m., 14 Sept. 1954, Stainton, Sykes & 
Williams y828. Arun Valley, Kasuwa Khola, 2,800 m., on tree in forest, 11 Sept. 
1956, Stainton 162 1. 

Distribution : Worldwide ; previously recorded from Nepal. 

PoLYPORUS zoNALis Berk, in Ann. & Mag. Nat. Hist. 10 : 375, t. 10 fig. 5 (1843). 

Nepal: Karelung, Madi Khola, 660 m., on rotten branch, 23 June 1954, Stainton, 

Sykes & Williams 3911. 



FUNGI OF RECENT NEPAL EXPEDITIONS 125 

Distribution: tropical and semi-tropical; Central and South America, Cuba, 
India, Indonesia, China, Australia. 
Resembles the type but is a little thicker. Spores globose and no cystidia. 

PoLYSTiCTUS AFFiNis (Blumc & Nccs) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, 
I : 75 (1851)- 

Polyporus affinis Blume & Nees in Nov. Act. Phys.-Med. Acad. Caes. Leop.-Car. 13 : 18, 

t. 4 (1826). 
Microporus affinis (Nees) Kuntze, Revis. Gen. PI. 3 (2) ; 495 (li 



Nepal: Arun Valley, Sashaya Khola, 660 m., on tree trunk in forest, 4 Sept. 
1956, Stainton 1578. Dharan Bazar, Terai Forest, south of Gopa Gurkha Camp, 
250 m., 27 Feb. 1962, Norkett 10224. 

Distribution : widespread in tropical and subtropical regions ; previously recorded 
from Nepal. 

PoLYSTiCTUS ciNNAMOMEUS (Gray) Sacc, Syll. Fung. 6 : 210 (1888). 

Strilia cinnamomea Gray, Nat. Arrang. Brit. PI. i : 645 (1821). 

Polyporus cinnamomeus (Gray) Fries, Epicrisis Syst. Mycol. : 468 (1838). — Overh., Polyp. 

U.S., Alaska & Canada : 386 (1953) • 
Coltricia cinnamomea (Gray) Murrill in Bull. Torrey Bot. CI. 31: 343 (1904). 

Nepal: Taglung, Kali Gandaki, 3,500 m., in wood, 22 Sept. 1954, Stainton, Sykes 
& Williams yggi. 
Distribution : Worldwide ; previously recorded from Nepal. 
Very close to P. perennis but distinguished by its more uniform and silkier cap. 

POLYSTICTUS HiRSUTUS (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 86 
(1851). 

Polyporus hirsutus Fries, Syst. Mycol. i : 367 (1821). 

Coriolus hirsutus (Fries) Qudl., Enchir. Fung. : 175 (1886). — Bourd. & Galz., Hymdnomyc. 
Fr. : 561 (1928). 

Nepal: Dhankuta Province, Milke Danda Forest, 260 m., 29 Nov. 1961, Norkett 
8307 A. 

Distribution : Worldwide. 

POLYSTICTUS PERULA (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 73 (1851). 

Polyporus perula Fries, Syst. Mycol. i : 349 (1821). 

Polyporus xanthopus Fries, Syst. Mycol. i : 350 (182 1). 

Polystiotus xanthopus (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 74 (1851). 

Microporus perula (Fries) Harlot in Bull. Soc. Mycol. Fr. 7 : 206 (1891). 

Nepal: Rupakot Tal, 800 m., on rotten branch, 5 May 1954, Stainton, Sykes 6- 
Williams 5233. Arun Valley, Hinwan Khola, 800 m., on rotten log, 4 Sept. 1956, 



126 FUNGI OF RECENT NEPAL EXPEDITIONS 

Stainton 1533. Sanghu, 3,000 m., on tree stump, 2 Oct. 1961, Norkett 5540. Chain- 
pur district, Tumlingtar, 600 m., 13 Dec. 1961, Norkett 8iog C. 

Distribution: Widespread in tropical and sub-tropical areas; previously recorded 
from Nepal. 

PoLYSTiCTUS SANGUINEUS (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 75 

(1851). 

Polyporus sanguineus Fries, Syst. My col. i : 371 (1821). 

Pycnoporus sanguineus (Fries) Murrill in Bull. Torrey Bot. CI. 31 : 421 (1904). 

Nepal: Midam Khola, Chisankhu, 660 m., 4 May 1954, Stainton, Sykes & Williams 
5218. Midam Khola, Karelung, 600 m., 23 June 1954, Stainton, Sykes & Williams 
Sgio. Kusma, 660 m., 2 Nov. 1954, Stainton, Sykes & Williams g2jo. Sanghu, 
3,000 m., 6 Oct. 1961, Norkett 5yog. 

Distribution : Mostly tropical and sub-tropical. 

The Nepal material is thin and smooth and conforms with P. sanguineus. P. 
cinnabarinus Fries, at one time considered to be a synonym, has been shown to be 
distinct on the basis of cultural interfertility tests (McKay in Mycologia, 51 : 465- 

73 (1959))- 

PoLYSTiCTUS suBAFFiNis Lloyd, Mycol. Not. 40 : 550, fig. 755 (1916). 

Nepal: Surauti Khola, 660 m., on dead bamboo, 12 Aug. 1954, Stainton, Sykes & 
Williams 686g. 

Distribution: Japan, Java, Madagascar. 

The present collection appears to agree completely with Lloyd's species as he 
figured it from Umemura's Japanese specimen, but it is doubtful whether this species 
is distinct from Polystictus affinis (Fries) Fries, 

POLYSTICTUS TABACiNus (Mont.) Sacc, Syll. Fung 6 : 280 (1888). 
Polyporus tabacinus Mont, in Ann. Sci. Nat., Ser. 2, Bot. 3 : 349 (1835). 

Nepal: Siklis, north of Pokhara, 2,100 m., on rotten tree trunk, 22 Apr. 1954, 
Stainton, Sykes & Williams 4gy4. Gurjagaon, 3,000 m., on dead tree, 25 Sept. 
1961, Norkett 5481. Milke Danda Forest, 3,000 m., on old dead tree, 16 Nov. 1961, 
Norkett 7i2y. 

Distribution: South America, Africa, India, East Indies, China, Australasia. 

9-10 pores per mm., setae subulate, dark brown. This species differs from 
P. iodinus in having smaller pores. 

Polystictus tephroleucus (Berk.) Sacc, Syll. Fung. 6 : 275 (1888). 

Trametes tephroleuca Berk, in Hook., Journ. Bot. 6 : 165 (1854). 

Coriolus tephroleucus (Berk.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. & Kavk. : 

492, fig. 126 (1953)- 
Coriolus favoliporus Pilat in Bull. Soc. Mycol. Fr. 52 : 313, t. 3 figs. 3-4 (1937). 
Trametes favolipora (Pildt) Pilat in Kav. & Pilat, Atlas Champ. 3 : 267, t. 182, fig. 105 

(1939)- y 



FUNGI OF RECENT NEPAL EXPEDITIONS 127 

Nepal: Lulo Khola, i6 Sept. 1952, Polunin, Sykes & Williams 3448. Near 
Lumsum, 2,300 m., on dead tree stump, 24 Oct. 1954, Stainton, Sykes & Williams 

9137- 

Distribution: Asia: Kazakstan, India and East Nepal. 

Fine specimens but the pores in Stainton, Sykes & Williams gisy are mostly 

discoloured owing to a mycelial growth over the hymenium. 

PoLYSTiCTUS VERSATiLis (Berk.) Fries in Nov. Act. Soc. Sci. Upsal., Ser. 3, i : 92 

(1851). 

Trametes versatilis Berk, in Lond. Journ. Bot. i : 150 (1842). 

Polyporus versatilis (Berk.) Romell in Bih. K. Svensk. Vet. -Akad. Handl. 26, (3, 16) : 35 

(1901). — Lloyd, Mycol. Not. 50 : 703, figs. 1049-50 (1917). — Overh., Polyp. U.S., 

Alaska & Canada: 325 (1953) • 

Nepal: Ghar Khola, 2,600 m., 3 May 1954, Stainton, Sykes & Williams 5444. 
Maikot, 2,600 m., on stump, 4 July 1954, Stainton, Sykes & Williams 3363. Tum- 
lingtar, Chainpur, 600 m., on old tree in ravine, 9 Dec. 1961, Norkett 8693. Dharan 
Bazar, Terai forest, south of Gopa Gurkha Camp, 230 m., 27 Feb. 1962, Norkett 
10232. 

Distribution: America, Madagascar, India, Malaya, Indonesia, Japan, China. 

PoLYSTiCTUS VERSICOLOR (Fries) Fries in Nov. Act. Soc. Sci. Upsal., Ser. 3, i : 86 

(1851). 

Polyporus versicolor Fries, Syst. Mycol. i : 368 (1821). 
^ Coriolus versicolor (Fries) Qu61., Enchir. Fung. : 175 (1886). 

Nepal: Ranipauwa, Kali Gandaki, 1,000 m., 3 Sept. 1954, Stainton, Sykes & 
Williams 7637. Lete, Kali Gandaki, 2,800 m., 17 Sept. 1954, Stainton, Sykes & 
Williams 78g3. Bakhri Kharka, Pokhara, 2,000 m., 25 Apr. 1954, Stainton, Sykes 
& Williams 5073. Arun Valley, Hatiar, 2,600 m., 21 Aug. 1956, Stainton 1412. 
Taplejung, above Sanghu, Milke Danda Forest, 3,000 m., 16 Nov. 1961, Norkett 
yi24 A. Mewa Khola, 2,700 m., 23 Jan. 1962, Norkett 9343 and 9343 A. 

Distribution: Worldwide; previously reported from Nepal. 

Daedalea unicolor Fries, Syst. Mycol. i : 336 (1821). — Overh., Polyp. U.S., 
Alaska & Canada : 125 (1953). 

Coriolus unicolor (Fries) Patouill., Ess. Tax. Fam. & Genr. Hym6nomyc. : 94 (1900). 

Nepal: Sanghu, 3,000 m., i Nov. 1961, Norkett 6y28. 
Distribution : America, Europe, North Africa, China, AustraUa. 

Trametes cervina (Schwein.) Bresad. in Ann. Mycol. i : 81 (1903). 

Boletus cervinus Schwein. in Schrift. Naturf. Ges. Leipz. i : 96 (1822). 

Coriolus cervinus (Schwein.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. & Kavk. : 

493, fig- 127 (1953)- 
Polyporus biformis sensu Berk, in Ann. & Mag. Nat. Hist. 3 : 392 (1839), non Klotzsch. 



128 FUNGI OF RECENT NEPAL EXPEDITIONS 

Nepal: UUeri, north of Kusma, Kali Gandaki, 2,600 m., on tree in wood, i Nov. 
1954, Stainton, Sykes & Williams Szy^. 

Distribution: Europe, Russia, India, Ceylon, China, Australia. 

For nomenclature of this fungus see note under Polyporus pargamenus (p. 123). 

Trametes gibbosa (Fries) Fries, Epicrisis Syst. Mycol. : 492 (1838). 

Daedalea gibbosa Fries, Syst. Mycol. i : 338 (1821). 

Pseudotrametes gibbosa (Fries) Bondartz. & Sing, in Ann. Mycol. 39 : 60 (1941). 

Nepal: Siklis, north of Pokhara, 21 Apr. 1954, Stainton, Sykes & Williams 4957. 
Distribution: Europe, Africa, India, China. 

Lenzites betulina (Fries) Fries, Epicrisis Syst. Mycol. : 405 (1838). 

Daedalea betulina Fries, Syst. Mycol. i : 333 (1821). 
Trametes betulina (Fries) Pildt in Kav. & PUat, Atlas Champ. Eur. 3 : 327, t. 220, fig. 142 
(1940). 

Nepal: Arun Valley, Hatiar, 2,600 m., on tree trunk, 21 Aug. 1956, Stainton 1412. 
Sikhs, north of Pokhara, 2,500 m., on rotten tree trunk, 19 Apr. 1954, Stainton, 
Sykes & Williams 4926. 

Distribution: Worldwide. 

Lenzites palisotii (Fries) Fries, Epicrisis Syst. Mycol. : 404 (1838). 

Daedalea palisotii Fries, Syst. Mycol. i : 335 (1821) " Palisoti ". 

Daedalea applanata Klotzsch in Linnaea 8 : 481 (1833). 

Lenzites repanda Fries, Epicrisis Syst. Mycol. : 404 (1838). 

Lenzites applanata (Klotzsch) Fries, Epicrisis Syst. Mycol. : 404 (1838). 

Nepal: Arun Valley, Sabhaya Khola, 800 m., on tree trunk, 3 Sept. 1956, Stainton 
i$y6. Dhankuta Province, Chainpur district, Tumlingtar, 600 m., on dead branch, 
13 Dec. 1961, Norkett 8iog A. 

Distribution: Widespread, especially in the southern hemisphere. 

Lenzites subferruginea Berk, in Hook., Journ. Bot. 6 : 134 (1854). 

Gloeophyllum subferrugineum (Berk.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. & 
Kavk. : 50 (i953)- 

Nepal: Lete, Kali Gandaki Valley, 2,600 m., on rotten tree trunk, 12 June 1954, 
Stainton, Sykes & Williams 3730; same locality, 8 July 1954, Stainton, Sykes & 
Williams i6jg. Taglung, Kali Gandaki, 27 Aug. 1954, Stainton, Sykes & Williams 

7494- 

Distribution: India, Nepal, Philippines, Japan. 

No spores were found in any of the gatherings. Stainton, Sykes & Williams 3730 

has a grey cap, 1639 has grey cap with wide brown margin, and in 7494 the cap is 

entirely brown. 



FUNGI OF RECENT NEPAL EXPEDITIONS 129 

THELEPHORACEAE 

CoRTiciUM CAERULEUM (Pers.) Fries, Epicrisis Syst. Mycol. : 562 (1838). 

Thelephora caerulea Pers., Mycol. Eur. i : 147 (1822). — Fries, Elench. Fung, i : 202 

(1828). 

Nepal: Ganesh Himal, Ankhu Khola, 2,800 m., on bark in broad-leaved forest, 
17 May 1962, Stainton 3731. 
Distribution: America, Europe, India, Australia, Japan. 

Hymenochaete mougeotii (Fries) Cooke in Grevillea 8 : 147 (1880). 
Thelephora mougeotii Fries, Elench. Fung, i : 188 (1828). 

Nepal: Annapurna Himal, on branches of Rhododendron campanulatum , 30 Aug. 
1954, Stainton, Sykes & Williams 6641. Rambrong, Lamjung Himal, on branch of 
Betula utilis, 7 July 1954, Stainton, Sykes & Williams 6202. Mewa Khola, 1,300 m., 
on dead wood, i Feb. 1962, Norkett gi82. Ganesh Himal, Ankhu Khola, 2,800 m., 
on bark in broad-leaved forest, 17 May 1962, Stainton 3732. Ganesh Himal, 
Mailung Khola, 4,000 m., 20 May 1962, Stainton 3744. 

Distribution: Europe, India, Nepal, Australia, New Zealand, China. 

Hymenochaete rheicolor (Mont.) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 151 
(1846). 

Stereum rheicolor Mont, in Ann. Sci. Nat., Ser. 2, Bot. 18 : 23 (1842). 
Stereum tenuissimum Berk, in Hook., Lond. Journ. Bot. 6 : 510 (1847). 
Hymenochaete sallei Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 333 (1868). 
Hymenochaete tenuissima Berk, apud Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 

333 (1868) worn. nud. 
Hymenochaete tenuissima (Berk.) Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : 

67 (1873). 
Stereum elegantissimum Spegazz. in An. Soc. Cient. Argent. 17 : 78 (i^ 



Nepal: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., 17 Oct. 1961, 
Norkett ^6g6 B; same locality, 15 Nov. 1961, Norkett yii2 C. 
Distribution: North and South America, Africa, West Indies, India. 

Hymenochaete rubiginosa (Fries) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 151 

(1846). 

Thelephora rubiginosa Fries, Syst. Mycol. i : 436 (1821). 

Nepal: Chainpur district, Tumlingtar, 600 m., on dead tree, 13 Dec. 1961, Norkett 
8iog D; same locahty, 16 Dec. 1961, Norkett 8567. 
Distribution : Worldwide. 



I30 FUNGI OF RECENT NEPAL EXPEDITIONS 

Hymenochaete tabacina (Fries) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 152 (1846). 

Thelephora tabacina Fries, Syst. Mycol. i : 437 (1821). 

Stereum tabacinum (Fries) Fries, Epicrisis Syst. Mycol. : 550 (1838). 

Nepal: Annapurna Himal, on branches of Rhododendron campanulafum, 3 Aug. 
1954, Stainton, Sykes & Williams 6642. Near Dogadi Khola, 4,300 m., on dead 
shrubs, II Aug. 1954, Stainton, Sykes & Williams 3818. 

Distribution: Worldwide. "^ 

LoPHARiA CRASSA (Lev.) Boidiu in Bull. Soc. Mycol. Fr. 74 : 479 (1958). 

Thelephora crassa L^v. in Ann. Sci. Nat., Ser. 3, Bot. 2 : 209 (1844). 

Stereum umbrinum Berk. & Curt, apud Berk, in Grevillea i : 164 (1873). 

Hymenochaete vinosa Cooke in Grevillea 8 : 149 (1880). 

Laxitextum crassum (L6v.) Lentz in U.S. Dept. Agric, Agric. Monogr. 24 : 20 (1955). 

Nepal: Dhankuta Province, Dhankuta, near Mahe, 1,300 m., 20 Sept. 1961, 
Norkett 5i5y C. Sombu, 1,600 m., 23 Sept. 1961, Norkett 528g. Dhankuta Pro- 
vince, Taplejung district, Sanghu, 3,000 m., 3 Oct. 1961, Norkett ^Sjg A; same 
locality, 2,000 m., 15 Nov. 1961, Norkett yii2 C. 

Distribution: America, Europe, Africa, India, Australia, New Zealand. 

Stereum ostrea (Fries) Fries, Epicrisis Syst. Mycol. : 547 (1838). 

Thelephora fasciata Schwein. in Schrift. Naturf. Ges. Leipz. i : 106 (1822). 

Thelephora ostrea Fries, Elench. Fung, i : 175 (1828). 

Thelephora versicolor vox. fasciata (Schwein.) Fries, loc. cit. 

Stereum fasciatum (Schwein.) Fries, Epicrisis Syst. Mycol. : 546 (1838). 

Nepal: Arun Valley, Khandbari, 2,300 m., on rotting log, 31 Aug. 1956, Stainton 
1472- 

Distribution: widespread; previously recorded from the Himalayas. 

Since Fries did not treat Thelephora fasciata Schwein. as a separate species in his 
Elenchus, which is part of the starting-point for the Fungi caeteri, the epithet of his 
T. ostrea must be adopted when the two names are regarded as synonyms. 

Stereum hirsutum (Fries) Gray, Nat. Arrang. Brit. PI. i : 653 (1821). 
Thelephora hirsuta Fries, Syst. Mycol. i : 439 (182 1). 

Nepal: Chipli, north of Pokhara, 3,600 m., 18 Apr. 1954, Stainton, Sykes &\ 
Williams 4885. Arun Valley, Hatiar, 2,600 m., on rotting log, 21 Aug. 1956,] 
Stainton 1410. Arun Valley, Sibrung, 27 Aug. 1956, Stainton 144J. Kasuwa Khola, j 
3,300 m., 12 Sept. 1956, Stainton 163^. Milke Danda Forest, 3,000 m., 16 Nov. 
1961, Norkett yi2$. Sanghu, 2,300 m., 23 Nov. 1961, Norkett 8ig6. 

Distribution : Worldwide. 



\ 



FUNGI OF RECENT NEPAL EXPEDITIONS 131 

Stereum roseo-carneum (Schwein.) Fries, Acta Soc. Sci. Upsala, Ser. 3, i : 112 
(1851). 

Thelephora roseo-carnea Schwein. in Schrift. Naturf. Ges. Leipz. i : 107 (1822). 
Laxitextum roseo-carneum (Schwein.) Lentz in U.S. Dept. Agric, Agric. Monogr. 24 : 22 
(1955)- 

Nepal: Dhankuta Province, Taplejung district, Sanghu 2,060 m., 15 Nov. 1961, 
Norkett yii2 B. 

Distribution: North and South America, Japan, China. 

The fructifications are pinkish buff, resupinate on twigs; the paraphyses have 
branching tips; spores 8 X 4-5 /*. 

Stereum sanguinolentum (Fries) Fries, Epicrisis Syst. Mycol. : 549 (1838). 

Thelephora sanguinolenta Fries, Syst. Mycol. i : 440 (1821). 

Nepal: Gurjakhani, 3,160 m. on small branches, 30 July 1954, Stainton, Sykes & 
Williams 36y8. 
Distribution: North America, Europe, South Africa, Australia, New Zealand. 

Stereum subpileatum Berk. & Curt, apud Berk, in Hook., Journ. Bot. i : 238 
(1849). — Lentz in U.S. Dept. Agric, Agric. Monogr. 24 : 36 (1955). 

Stereum insigne Bresad. in Nuovo Giorn. Bot. Ital. 23 : 158 (1891). 

Xylobolus subpileatus (Berk. & Curt.) Boidin in Rev. de Mycol. 23 : 336 (1958).- — Lentz 
in Sydowia 14 : 118 (i960). 

Nepal: Bakhri Kharka, north of Bokhara, on rotten tree trunk, 24 Apr. 1954, 
Stainton, Sykes & Williams 50 ^g. Milke Danda Forest, Dhankuta Province, 2,800 
m., 2 Nov. 1961, Norkett 6806 A ; same area, 3,000 m., 16 Nov. 1961, Norkett 7124. 

Distribution: America, Europe, India, Indonesia, China, Japan. 

Lentz and Boidin should be consulted for modern interpretations of this aggregate 
species. 

Stereum sp. cf. sulcatum Burt apud Peck in New York State Mus. Annu. Rep. 
54, I, App. I : 154 (1901). 

Nepal: South of Gurjakhani, 3,600 m., on tree in Abies forest, 16 Aug. 1954, 
Stainton, Sykes & Williams 3868. 

Large robust sporophores superficially hke those of a large S. princeps (Jungh.) 
Lev. but having a tomentose sulcate pileus, coarse, more or less parallel skeletal 
hyphae intermingled with generative hyphae, and no acanthophyses as such, but 
merely some shghtly granular cystidial hyphae. The spores, irregularly globose, 
smooth or very faintly punctate and amyloid, 4-5-6 /i diameter, resemble those of 
S. sulcatum Burt and those of 5. taxodii Lentz & McKay (Mycologia 52 : 262 (i960)), 
two species recently transferred by H. L. Gross to his genus Echinodontium (Myco- 
path. & Mycol. Appl. 24 : 8, 11 (1964)). The Nepal fungus however differs in the 



132 FUNGI OF RECENT NEPAL EXPEDITIONS 

absence of large encrusted cystidia and in the possession of large flabelliform reflexed 
pilei. In view of the large and conspicuous fructifications it would seem unlikely 
that this fungus has not been recorded previously. I therefore defer describing it 
as new to science. 

Thelephora caryophyllaea Fries, Syst. Mycol. i : 430 (1821). 

Phylacteria caryophyllea Patouill., Hym. Eur. : 154 (1887), nom. nud. 
Phylacteria caryophyllea (Fries) Patouill. ex Bourd. & Maire in Bull. Soc. Mycol. Fr. 36 : 
76 (1920). 

Nepal: Taplejung district, above Sanghu, 2,000 m., amongst moss on earth, 
12 Oct. 1961, Norkett 5g2y A. 
Distribution: Worldwide. 

Vararia rhodospora (Wakef.) G. H. Cunn. in Proc. Linn. Soc. N.S.W. 77 : 291 
(1953); in N.Z. Dept. Sci. Industr. Res. Bull. 145 : 100 (1963). 

Stereum duriusculum sensu Bresad. in Ann. Mycol. 6 : 43 (1908), non Berk. & Broome. 
Asterostromella rhodospora Wakef. in Kew Bull. 1915 : 372 (1915). — Banergee in Journ. 

Ind. Bot. Soc. 14 : 45 (1935). 
Asterostromella dura Bourd. & Galz. apud Bourd. & Maire in Bull. Soc. Mycol. Fr. 36 : 

74 (1920). 
Dichostereum durum (Bourd. & Galz.) Pildt in Ann. Mycol. 24 : 223 (1926). 

Nepal: Sanghu, 820 m., 9 Nov. 1961, Norkett 7324. 

Distribution: America, Europe, Africa, India, Japan, Australia, New Zealand. 

The species was recorded from India by Banergee. The Nepal collection con- 
sists of tough, resupinate thalli covering earth beneath tree roots, about 2-3 mm. 
thick, mid-brown, of stratose context. Only a few basidia were observed embedded 
in the dichophysoid paraphyses ; spores globose, straw-coloured, echinulate, 5-6 /i 
diameter; context hyphae brown, dendroidly and dichotomously branched, most 
markedly and densely in the hymenial layer, where they form the brown dendro- 
physes and dichophyses; intermingled with them are finer, readily stained hyphae. 
Rogers & Jackson (Farlowia i : 309 (1943)), treat Dichostereum durum as a synonym 
of Vararia pallescens (Schwein.) Rog. & Jacks. Type material of Thelephora pal- 
lescens Schwein. in the B.M. Herbarium differs in several particulars, notably its 
finer context, and would appear to be not merely a different growth-form, as these 
authors suggested, but a distinct species. 

CLAVARIACEAE 

Clavulina mussooriensis Corner, Thind & Dev in Trans. Brit. Mycol. Soc. 41 : 
204 t. 8 fig. 3, text-fig. I (1958). 

Nepal: Near Gurjakhani, 2,800 m., among grass on open slope, 28 July 1954, 
Stainton, Sykes & Williams 36^0. 
Distribution: India. 

(E.J.H.C.) 



FUNGI OF RECENT NEPAL EXPEDITIONS 133 

Clavulina alta Corner, sp. nov. 

Receptacula ad ii cm. alta, alba, sicco luride flava; stipite 2-6 cm. x 3-8 mm., 
bene evoluto; ramulis inferioribus polychotomis v. applanato-multifidis, superior- 
ibus I mm. latis dichotomis v. cristatis, axillis inferioribus 3-6 mm. latis. Sporae 
9-5-14 X 7-5-9 /*, subglobosae, lacrymiformes v. pyriformes, apiculo i ju, longo. 
Basidia 6-$-y-5 /li lata, bispora. Hymenium incrassatum; cystidiis nullis; hyphis 
subhymenialibus 6-17 /i latis, fere pseudoparenchymaticis. Hyphae 3-12 jii latae, 
fibulatae, tenue tunicatae, cellulis potius brevibus. 

Nepal: Chimgaon, north of Tukucha, Kali Gandaki, 3,500 m., on ground beneath 
conifers, 14 Sept. 1954, Stainton, Sykes & Williams y82^. (Herb. Mus. Brit, holotype). 
This resembles C. rugosa (Fries) Schroet. in the large spores and wide sub- 
hymenial hyphae, and C. cristata var. coralloides Corner in the form of the fruit-body. 
I have not seen such a distinct form before, and the spores are constantly rather 
narrow for their length. 

(E.J.H.C.) 

Lentaria macrospora Corner, sp. nov. 

Receptacula ad 10 cm. alta, gregaria v. caespitosa, carneoflavida ; stipite ad 
25 X 2-4 mm., axillas inferiores polychotomas versus plus minus dilatato; ramulis 
superioribus teretibus dichotomis strictis ascendentibus, 1-2 mm. latis. Sporae 
20-30 X 3 7-5 '5 /^, hyalinae, cylindricae, v. subclavatae, saepe curvatae v. sig- 
moideae, et allantiformes, tenue tunicatae, hand amyloideae. Basidia ad 45 X 
9-10-5 /i; sterigmatibus 2-4, 7-8 ju, longis. Hymenium incrassatum; cystidiis nullis. 
Hyphae 2-5-7 M' latae, fibulatae, tunicis ad 0-5 /i vix incrassatis; in mycelio 2-5-4 /* 
latae, fibulatae, monomiticae, tunicis tenacibus sed vix incrassatis, passim partibus 
ampuUiformibus ad 15 jti latis inflatae, crystallis sphaeroideis 2-7 /i latis inter hyphas 
numerosis. 

Nepal: Tamur Valley, Ghunsa, east of Walungchung Gola, 4,300 m., on ground 
under conifers, 27 July 1956, Stainton 1145. (Herb. Mus. Brit, holotype). 

This resembles in shape and colour the common tropical L. surculus (Berk.) 
Corner but the spores are much longer, the hyphal walls are scarcely thickened, and 
the habitat seems to be humicolous. Many basidia had 1-3 long spiculiform sterig- 
mata, but they may have been abnormal and formed after collection. 

(E.J.H.C.) 

Ramaria aff. BOTRYTOiDES (Pcck) Comer, Monogr. Clavaria, Ann. of Bot., Mem. 

I : 562 (1950). 

Clavaria hotrytoides Peck in N.Y. State Mus., Mus. Bull. 94 : 49, t. 93 figs. 5-7 (1905)- 

Nepal: Taglung, south of Tukucha, Kali Gandaki, 3,300 m., on ground beneath 
trees, 11 July 1954, Stainton, Sykes & Williams i6gi. 

Distribution (of Ramaria hotrytoides) : America, southern parts of Australia and 
Tasmania, Japan ; when the species is interpreted in a wide sense. 

(E.J.H.C.) 



134 FUNGI OF RECENT NEPAL EXPEDITIONS 

Ramaria suecica (Fries) Donk in Med, Bot. Mus. Herb. Rijks Univ. Utrecht 9 : 
105 (1933). — ^Corner, Monogr. Clavaria, Ann. of Bot., Mem. i : 629 (1950). 

Clavaria suecica Fries, Syst. Mycol. i : 469 (182 1). 

Clavariella suecica (Fries) Karst. in Bidr. Kann. Finl. Natur. & Folk 37 : 187 (1882). 

Nepal: South of Gurjakhani, 3,300 m., on damp shady forest bank, 18 Aug. 
1954, Stainton, Sykes & Williams 3go2. 
Distribution: Europe, China, Canada, Northern U.S.A. 

(E.J.H.C.) 

Ramaria flaccida (Fries) Ricken, Vadem, Pilzfr. : 254 (1918). 

Clavaria flaccida Fries, Syst. Mycol. i : 471 (1821). 

Nepal: Taglung, south of Tukucha, KaU Gandaki, 3,300 m., in pine wood, 22 
Sept. 1954, Stainton, Sykes & Williams 7971. 

Distribution: America, Europe, South Africa, Austraha, China, Japan. 

(E.J.H.C.) 

Ramaria obtusissima (Peck) Corner, Monogr. Clavaria, Ann. of Bot., Mem. i : 609 

(1950). 

Clavaria obtusissima Peck in N.Y. State Mus., Mus. Bull. 167 : 39 (1913). 

Nepal: Taglung, south of Tukucha, Kali Gandkai, 3,000 m., beneath conifers, 
12 July 1954, Stainton, Sykes & Williams lygo. 
Distribution: U.S.A. and Canada. 
The pink form is represented. 

(E.J.H.C.) 

Ramaria subaurantiaca Corner apud Balf. -Browne in Bull. Brit. Mus. (Nat. Hist.), 

Bot. I : 200 (1955). 

Nepal: North of Barse, 4,000 m., on Ahies stump, 14 Aug. 1954, Stainton, Sykes 
& Williams 3851. 

Distribution: Tibet. 

Spores 10-5-15 X 5-6 {I, rather coarsely subverrucose. No clamps. 

(E.J.H.C.) 

cyphellaceae 
Chrotnocyphella bryophyticola Balfour-Browne, sp. nov. 

Fungus cupulatus, cupulis sessilis, levis, griseo-albis, 0-5 mm. diam. ; contextis 
tenuis, mollis, 15/* latis ex hyphis elongatis efformatis; hymenio levo, brunneo;} 
basidiis cylindricis, 15-16 X 4-5 jn] sporis globosis, 5-7 ju, diam., brunneis, punctatis. 

Nepal: Sanghu, 1,800 m., on moss, Pterobryopsis, and on an intermingled liver-] 
wort, on shady rock, 8 Nov. 1961, Norkett 7292. (Herb. Mus. Brit, holotype). 



I FUNGI OF RECENT NEPAL EXPEDITIONS 135 

This fungus is not unlike Cyphella chromospora Patouill. (Tab. Anal. Fung, i : 19, 
'. 32 (1883)), but the spores are larger. It differs from Chromocyphella burtii 
idge Cooke (Sydowia, Beiheft 4 : 137 (1961)), in its smaller basidia and its definitely 
punctate spores. 

This appears to be the first record of a " Cyphella " in the Himalayan region. 

SCLERODERMATEACEAE 

Scleroderma aurantium Pers., Syn. Meth. Fung, : 153 (1801). 

Nepal: Mathand, near Pokhara, 1,160 m., on shady bank, 22 June 1954, Stainton, 
Sykes & Williams 5851. 
Distribution : Worldwide. 
Spores 8-10 [i, reticulated. 

LYCOPERDACEAE 

BoviSTA sp. cf. BOViSTOiDES (Cooke & Massee) Ahmad, Gasteromycetes W. Pakistan, 
Publ. Dept. Bot. Univ. Panjab 11 : 16 (1952). 

Mycenastrum bovistoides Cooke & Massee in Grevillea 16 : 26 (1887). 

Nepal: Above Dogadi Khola, amongst short grass on exposed slope, 21 June 
1954, Stainton, Sykes & Williams 3211. 

Distribution (of Bovista bovistoides) : India. 

The Nepal collection appears to be very close to this species but has very slightly 
warted spores, whereas Ahmad described them as smooth. The capillitium threads 
agree in being chestnut brown and unpitted. 

BoviSTA ECHINELLA Patouill. in Bull. Soc. Mycol. Fr. 7 : 165 (1891). 

Bovistella echinella (Patouill.) Lloyd, Mycol. Notes 23 : 286, t. 89 figs. 1-2 (1906). 
Lycoperdon echinella (Patouill.) Ahmad in Journ. Ind. Bot. Soc. 20 : 138 (1941). 

Nepal: Chainpur district, Tumlingtar, 600 m., on earth near base of cliff of 
Sabhaya River, 9 Dec. 1961, Norkett 8682 A. 

Distribution : North and South America, Jamaica, Europe, Pakistan. 

The spores are smooth, not echinulate as described by Patouillard, but show 
" lines " beneath the outer membrane which at a certain focus appear like spines; 
the pedicels are mostly 6 /* long. 

Geastrum fimbriatum (Fries) E. Fisch. in Engl. & Prantl, Nat. Pflanzenfam., 2 
Aufl., 7a : 73 (1933)- 

Geaster fimbriatus Fries, Syst. Mycol. 3 : 16 (1829). 

Nepal: Tukucha, Kali Gandaki, 3,600 m., 12 Oct. 1954, Stainton, Sykes & 
Williams 8212. 
Distribution: America, Europe, Africa, India, Australia. 



136 FUNGI OF RECENT NEPAL EXPEDITIONS 

Geastrum hariotii (Lloyd) E, Fisch. in Engl. & Prantl, Nat. Pfianzenfam., 2 Aufl., 
7a : 73 (1933). 

Geaster hariotii Lloyd, Mycol. Not. 25 : 311, t. 99 figs. 7-9 (1907). 

Nepal: Tumlingtar, Sabhaya Khola, 600 m., on earth beneath bamboo, 20 Dec. 
1961, Norkett goo2. 

Distribution: South America, West Indies, Europe, Ceylon, Australia (according 
to Cunningham, v. below). 

The eight specimens from Nepal resemble the descriptions of Geastrum hariotii 
very closely in being non-hygroscopic, in having a sessile endoperidium, dark, 
sulcate peristome and small, 3-3-5 ju, diam., minutely verrucose spores. But the 
surface of the endoperidium is furfuraceous or granular rather than pitted as des- 
cribed by some authors (Cunningham, Gasteromycetes Austral, and N.Z. : 165 
(1942)). 

Lycoperdon pyriforme Pers., Syn. Meth. Fung. : 148 (1801). 

Nepal: Tukucha, Kah Gandaki, 3,500 m., 12 Sept. 1954, Stainton, Sykes & 
Williams y8o6. 

Distribution: Worldwide. 

FUNGI IMPERFECTI 

SPHAEROPSIDACEAE 

CoNiOTHYKiNA AGAVES (Durieu & Mout.) Petr. & Syd. in Beih. Rep. Spec. Nov. 
Regn. Veg. 42 : 322 (1927). 

Phoma agaves Durieu & Mont, in Mont., Syll. PI. Crypt.: 271 (1856). 
Coniothyrium agaves (Durieu & Mont.) Sacc, Syll. Fung. 3 : 318 (1884). 

Nepal: Sanghu, 2,000 m., on Agave, 17 Nov. 1961, Norkett 7154; Sombu, 1,600 
m., on Agave, 23 Sept. 1961, Norkett 5313. 

Distribution: America, South Europe, Africa, India. 

nectrioidaceae 
AscHERSONiA sp. cf. viRiDANS (Berk. & Curt.) Patouill. in Bull. Soc. Mycol. Fr. 
7 : 48 (1891)- 

Hypocrea viridans Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 376 (1868). 
Aschersonia disciformis Patouill. in Bull. Soc. Mycol. Fr. 8 : 136 (1892). 
Aschersonia viridula Sacc. in Ann. Mycol. 11 : 547 (191 3). 

Nepal: Phewa Tal, 800 m., on leaves of Castanopsis sp., 8 May 1954, Stainton, 
Sykes & Williams $2^2 [a). 

Distribution (of A . viridans) : Central America (Trinidad, Vera Cruz, Cuba, Ecua- 
dor), Brazil, Mexico. 

The Nepal fungus was growing on white fly. It differs from previous accounts in 
the greater number of pycnidia to each stromatic cushion, which is completely dotted 
over with the greenish ostioles; conidia 12-16 X 1-5-2 /*. 



I 



I 



FUNGI OF RECENT NEPAL EXPEDITIONS 137 



Another Aschersonia also was collected on Castanopsis leaves; this appears to be 
close to A. caespiticia Syd. (in Engl., Bot. Jahrb. 54 : 260 (1916)), but differs in 
the rough surface of the tubercles and its small basal cushions. Tilhar, 3 Nov. 
1954, Stainton, Sykes & Williams 92^1 {a). 

LEPTOSTROMATACEAE 

Melasmia salicina Tulasne frat., Sel. Fung. Carp. 3 : 119, t. 15 figs. 15-17 (1865). 

Nepal: East of Chalike Pahar, 4,160 m., on Salix sp., 25 Sept. ig^^, Stainton, 
Sykes & Williams 458y. 

Distribution: not previously collected in this part of the world but its perfect 
state, Rhytisma salicinum Fries, has been recorded from the Punjab. 

MELANCONIACEAE 

Mastigonetron americanum (Mont.) Balfour-Browne, comb. nov. 

Pestalotia americana Mont, in Gay, Hist. Chile, Bot. 7 : 481 (1850). — Guba, Monogr. 

Monochaetia & Pestalotia : 268 (1961). 
Seiridium Hquidambaris Berk. & Curt, apud Berk, in Grevillea 2 : 154 (1874). 
Mastigonetron fuscum Klebahn in Myc. Centralbl. 4 : 18, fig. 37-38 (1914). 
Monochaetia Hquidambaris (Berk. & Curt.) Guba, loc. cit. (1961), nom. synon. 

Nepal: Taplejung district, Sanghu, Tamrang Khola, 2,000 m., on dead twigs, 
19 Oct. 1961, Norkett 6319 A. 

Distribution: North and South America. 

The acervuli, about 0-5 mm. in diameter, are scattered over the twigs and resemble 
lenticels in appearance. The conidia, 20-27 X 9-10-5 /i, are dark brown, unicellular, 
ovoid or elhpsoid, each with a hyaline apical appendage 30-40 x 2 jn, and a pedicel 
12 X 2 ju,, approximately. 

The fungus agrees exactly with Klebahn's species and apparently also with the 
type of P. americana (in spite of Montague's description of the conidia as biseptate) 
since Guba, loc. cit., states that " Montague's drawings of the fungus and my study 
show dark colored i-celled elhpsoid conidia ", The drawing mentioned is apparently 
unpubhshed. 

In view of its unicellular conidia the species cannot be included in either Mono- 
chaetia or Pestalotia. As for Seiridium Nees & Henry, (Syst. Pilze : 18, t. 3 (1837)), 
arguments can be brought forward for retaining it as an earlier name for Mastigone- 
tron if one regards as accurate the elder Nees's (Syst. Pilze & Schwamme : 22 (1816), 
t. I fig. 19 (1817)) description and drawing of unicellular appendaged conidia. This 
non-septate condition of the conidia was accepted by Berkeley & Curtis when they 
selected this genus for the fungus on Liquidambar. Alternatively, Seiridium can 
be regarded as an older name for Monochaetia if Fries's (Syst. Mycol. 3 : 473 (1832)) 
statement that his examination of Nees's material showed multiseptate conidia is 
taken as correct. The original collection appears to be lost but other collections 
since then and reputed to be the same species, i.e. S. marginatum, are invariably 



138 FUNGI OF RECENT NEPAL EXPEDITIONS 

described as having a brown septate conidium with a transparent apical seta. The 
probable explanation for this discrepancy is that Nees chanced to examine and 
illustrate the unripe fungus, i.e. before the conidia became septate, while they were 
still unicellular, spindle-shaped and contained grey granular protoplasm. Fries, on 
the other hand, and all subsequent workers have described the mature fungus, which 
represents what is now generally regarded as typical Monochaetia. Taking this latter 
view, or better still treating Seiridium as a nomen confusum, this generic name cannot 
be used in place of Mastigonetron and therefore the correct name for the Nepal 
fungus becomes Mastigonetron americanum, as cited above. 



STILBACEAE 

Arthobotryum nilgirense Subram. in Proc. Ind. Acad. Sci., Sect. B, 46 : 324, 
fig. I (1957). 
Nepal: Chainpur, Tumlingtar, 660 m., on old bamboos, 15 Dec. 1961, Norkett 



Distribution: Previous record and original description from bamboo, Sim's Park. 
Coonoor, Madras, 1956. 

Podosporium himalensis Balfour-Browne, sp. nov. (Fig. 4). 

Synnemata atrobrunnea numerosa et effusa ex hyphis parallelibus juxtapositis 
efformata. Conidia atrobrunnea, 1-12 septata, obclavata, leniter curvata et irregu- 
lariter disposita, 9-15 x 16-100 /i. 

Nepal: Lamjung Himal, 4,500 m., black woolly growth on branches of Rhododen- 
dron campanulatum, 14 July 1954, Stainton, Sykes & Williams 6341. (Herb. Mus. 
Brit., holotype). 

The general form of the synnemata is similar to that described by Subramanian 
(Journ. Ind. Bot. Soc. 35 : 73 (1956)) for his Prathoda saparva; the conidia however' 
are different and the conidiophores hardly distinct from the hyphae and not geni- 
culate. The Nepal material resembles an extremely luxuriant form of Podosporium 
rigidum Schwein., originally described from Carolina. The hyphae are hormiscium- 
like and very probably any portion breaking away can regenerate fresh growth, 
independently of conidial reproduction. 

Stilbum cinnabarrinum Mont, in Ann. Sci. Nat., Ser. 2, Bot. 8 : 360 (1837). 
Stilbum lateritium Berk, in Ann. Nat. Hist. 4 : 291, pi. 8 fig. 2 (1840). 

Nepal: Tumlingtar, by shore of Sabhaya Khola, 600 m., 11 Dec. 1961, Norketi 
8107 C. 

Distribution: North and South America, Cuba, Dominica, Africa, Nepal, India,] 
Ceylon, Australia. 

Pleonectria pseudotrichia (Schwein.) Wollenw. is its perfect form. The present 
collection occurs on unnamed bark and is very sparing. 



FUNGI OF RECENT NEPAL EXPEDITIONS 



139 





Fig. 4. Podosporium himalensis Balfour-Browne. a, general aspect; b, conidia and 
conidiophorous hyphae. Stainton, Sykes &• Williams 6341. 



Stilbum inconspicuum Currey in Trans. Linn. Soc. Lond., Ser. 2, Bot, i : 129 
(1876). 

Stilbum kurzianum Cooke in Grevillea 16 : 71 (1888). 

Nepal: Sanghu, 2,000 m., 17 Nov. 1961, Norkett 7153. 
Distribution: India. 

On dead twigs of Rosa sp. Synnemata 3-4 mm. tall, pale orange soon becoming 
cinereous; conidia rod-shaped 7-9 X 3 /*. 



I40 FUNGI OF KECENT NEPAL EXPEDITIONS 

TUBERCULARIACEAE 

Epicoccum andropogonis (Rabenh.) Schol-Schwarz in Trans. Brit. Mycol. Soc. 
42 : 171, t. 9 fig. II (1959). 

Cerebella andropogonis Rabenh. in Bot. Zeit. q : 669 (1851). 

Nepal: Bhadauri, near Pokhara, 2,000 m., on inflorescence of a grass, i Nov. 
1954, Stainton, Sykes & Williams 8326. 

Distribution : Worldwide but chiefly in the tropics and subtropics. 

This genus has been revised recently by Langdon (Mycol. Commonw. Mycol. Inst. 
Papers, 61 : 1-18 (1955) under the name Cerebella), and by Schol-Schwarz (torn. cit. : 
149-173)- 

Epicoccum purpurascens Schlechtend., Fl. Berol. 2 : 136 (1824). — Link in L., Sp. 
PL, Ed. 4, 6 (2) : 108 (1825). 

Epicoccum nigrum Link, loc. cit. — Schol-Schwarz in Trans. Brit. Mycol. Soc. 42 : 170 
(1959). 

Nepal: Argam, near Pokhara, 830 m., on leaves and stems of Acrocephalus indicus 
(Burm.) O. Kuntze, 10 Sept. 1954, Stainton, Sykes & Williams 7146. 
Distribution : Worldwide. 

SPECIAL LITERATURE 

Balfour-Browne, F. Tv. (1955). Some Himalayan Fungi. Bull. Brit. Mus. {Nat. Hist.), 

Bot. I : 189-218. 
Berkeley, M. J. (1850). Decades of Fungi. Hook., Journ. Bot. 2 : 42-51, 76-88, 106-112. 

(1851). Decades of Fungi. Hook., Journ. Bot. 3 : 39-49, 77-84, 167-172, 200-206. 

(1852). Decades of Fungi. Hook., Journ. Bot. 4 : 97-107, 130-142. 

(1854). Decades of Fungi. Hook., Journ. Bot. 6 : 129-143, 161-174, 204-212, 225-235. 

Bondartzev, a. S. (1953). Trutov'ie grib'l Europe! skoi chasti S.S.S.R. i Kavkaza. [Poly- 

poraceae of the European part of the U.S.S.R. and the Caucasus.] Leningrad. 
Butler, E. J. & Bisby, G. R. (i960). Fungi of India. 2nd ed., revised by R. S. Vasudeva. 

New Delhi. 
Corner, E. J. H. (1932). The fruit-body oi Polystictus xanthopus. Ann. of Bot. 46 : 71-111. 

(1953)- The construction of Polypores. L Introduction. Phytomorphology 3 : 152-167. 

Cunningham, G. H. (1947). Notes on the classification of the Polyporaceae. N.Z. Journ. 

Sci. &' Technol. 28, A : 238-251. 

(1947-1950). New Zealand Polyporaceae. N.Z. Dept. Sci. Industr. Res. Bull. 72-83. 

(1965). Polyporaceae of New Zealand. N.Z. Dept. Sci. Industr. Res. Bull. 164. 

DoNK, M. A. (1933). Revision der niederlandischen Homobasidiomycetae — Aphyllophora- 

ceae. II. Med. Bot. Mus. Herb. Rijks Univ. Utrecht 9 : 1-278. 

(i960). The generic names proposed for Polyporaceae. Persoonia i : 173-302. 

Imazeki, R. (1943). [Genera of Polyporaceae of Nippon.] In Japanese. Bull. Tokyo Sci. 

Mus. 6, i-iii. 
KoTLABA, F. & PouzAR, Z. (i957). Poznamky k tfidSni evropskych choro.su. [Notes on 

the classification of European pore Fungi.] Ceskd Mykol. 11 : 152-170. 
Nobles, M. (1958). Cultural characters as a guide to the taxonomy and phylogeny of the 

Polyporaceae. Canad. Journ. Bot. 36 : 883-926. 



I 



FUNGI OF RECENT NEPAL EXPEDITIONS 141 

OvERHOLTS, L. O. (1953). The Polyporaceae of the United States, Alaska and Canada. Ann 

Arbor, Mich. 
PiLAT, A. (1936-1942). Polyporaceae. In Kavina, C. cS: Pilat, A., Atlas des Champignons 

de I' Europe, 3. Praha. 
Pinto-Lopez, J. (1952). " Polyporaceae " Contribui9ao para a sua biotaxonomia. Mem. Sac. 

Broter. 8 : 5-1 91. 
Sheals, J. G. & Inglis, W. G. (1965). The British Museum (Natural History) expedition to 

east Nepal 1961-62. Introduction and lists of localities. Bull. Brit. Mus. {Nat. Hist.), 

Zool. 12 : 95-114. 
Singer, R. (1962). The Agaricales in modern taxonomy. 2nd ed. Weinheini. 
Texieira, a. R. (1962). The taxonomy of the Polyporaceae. Biol. Rev. Cambr. Phil. Soc. 

37 : 51-81. 




\ 







:.''. ■, 



'■'{ I 



-■.''■V 









V -; ■ ■, . 



-?i>; 'v'\?^'t^' I 



&.^H^/-^.iV;;-:'V>.v> ■■V'-; ;A;::V^^., '..^n >>■ •■'.- ' ' ' 



;■■ '■■y ']y. 'Ur '/■'■> •:^-^V»ii^'-"v 




PRINTED IN GREAT BRITAIN 
BY ADLARD & SON LIMITED 
BARTHOLOMEW PRESS, DORKING 



f) : B 



A SYNOPSIS OF 
JAMAICAN MYRSINACEAE 



W. T. STEARN 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 4 

LONDON: 1969 



A SYNOPSIS OF 
JAMAICAN MYRSINACEAE 



BY 

WILLIAM THOMAS STEARN 



^M 



Pp. 143-173; 25 Text-figures; Plates 6-13 




BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY VoL 4 No. 4 

LONDON: 1969 



THE BULLETIN OF THE BRITISH MUSEUM 

(NATURAL HISTORY), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 4 of the Botany series. 



World List abbreviation 
Bull. Br. Mus. nat. Hist. (Bot.) 



Trustees of the British Museum (Natural History) 1969 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 30 May, 1969 Price £1 4s. 



I 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 

By WILLIAM T. STEARN 
Summary 

The family Myrsinaceae is represented in Jamaica by three genera, Ardisia, Wallenia and 
Myrsine (including Rapanea) . Recognition of the species presents especial difficulty in Wallenia, 
which is dioecious ; some taxa are known only from specimens with male flowers or female flowers 
or fruits, i.e. in only one of three states needed for adequate discrimination ; few are known in all 
three; 15 species, two occurring also in Cuba, the others endemic, are listed here. Ardisia is 
represented by eight species, one [A. solanacea) introduced and naturalized; Myrsine by two, 
which Mez placed in Rapanea. Keys, synonymy and statements of distribution are given. 
Ardisia brittonii and A. byrsonimae are described as new; A. rosea Urban (1908), non A. rosea 
King & Gamble (1905), is renamed A. urbanii. 

Introduction 

Students of the family Myrsinaceae tend to be censorious of the work of Carl Mez 
(1864-1944) on this difficult and to most botanists unattractive family, wherein, as 
C. L. Lundell (1966a) has remarked, " most generic characters are weak and those 
separating species are often weaker." Nevertheless, they must needs be grateful 
to Mez for his comprehensive survey of its West Indian members published in Urban's 
Symbolae Antillanae 2 : 389-433 (1901) and for his world monograph in Engler's 
Das Pflanzenreich IV. 236 (1902), in which he provided a helpful basic classification, 
co-ordinated the extensive literature, and enumerated and described the material 
then available. The subsequent great increase of specimens in herbaria and a more 
critical attention to matters of nomenclature and typification than prevailed then 
have, however, made necessary many alterations and some drastic revisions of Mez's 
treatment; moreover, some characters emphasized by him in his keys have been 
found impracticable. E. Y. Hosaka, for example, in his account of the Hawaiian 
species of Myrsine sensu lato (in Occ. Pap. B.P. Bishop Mus. 16 : 25-76 (1940)) 
concluded that Rapanea, Suttonia and Myrsine, kept as separate genera by Mez, 
overlap so much in their characters that they should be combined in one genus, 
Myrsine, as indeed they were by Bentham & Hooker in 1873 ; H. C. D. de Wit (in 
Bull. Jard. Bot. Bruxelles 27 : 233-242 (1957)) has united Heherdenia with Ardisia. 
On the other hand, Lundell (in W^rightia 3 : 88 (1963)) has raised Mez's Ardisia sect. 
Synardisia to generic rank as Synardisia. A few examples may suffice to indicate 
the nomenclatural rectifications needed. The name Ardisia coriacea Swartz (1788), 
based on Jamaican material, was misapplied by Mez to a species which does not 
occur in Jamaica, i.e. A. ohovata Ham. {A. guadalupensis Griseb.). The name 
Ardisia sanguinolenta Blume (1825) is the correct name for A. zollingeri A. DC. 
(1844): A. wallichii A. DC is the correct name for A. sanguinolenta Wall, ex Mez 
(1902). Merrill has called attention (in Lingnan Sci. Journ. 11 : 50 (1932) and in 
Contrib. Arnold Arb. 8 : 131 (1934)) to Mez's misinterpretation of A. humilis Vahl 
(1794), which, according to Merrill, is the same as A. hainanensis Mez (1902), while 

BOT. 4, 4 8 



146 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

A. humilis sensu Mez is A. elliptica Thunb. (1795) (syn, A. obovata Blume (Dec. 
1825-March 1826), non A. obovata Ham. (Nov. 1825)). As regards the East Asiatic 
Myrsinaceae, a much needed nomenclatural and taxonomic overhaul has been made 
by E. H. Walker (in Philipp. Journ. Sci. 73 : 1-258 (1940), with supplementary 
notes in Journ. Arnold Arb. 23 : 344-355 (1942), in Bot. Mag. Tokyo 67 : 105-111, 
155-162, 203-213, 247-255 (1954) and in Quart. Journ. Taiwan Mus. 12 : 161-194 
(1959)). Unfortunately, there exists no comprehensive survey of the American 
Myrsinaceae later than Mez's of 1902, although many new species have been published 
since by Urban, Standley and C. L. Lundell, who has also revised the genus Parathesis 
(Lundell, 19666) and the Guatemalan members of the family (Lundell, 1966a). The 
author of a local Flora is accordingly obliged to make a study more or less independent 
of Mez's work and to extend his enquiries beyond his own geographical area. The 
following synopsis of Jamaican Myrsinaceae is a by-product of preparing a more 
detailed account for the Flora of Jamaica ; descriptions and particulars of localities 
to be given there in detail are either summarized below or omitted. 

Mez in 1901 based his account of the Myrsinaceae occurring in Jamaica on 14 wild 
and two cultivated gatherings of Ardisia, 15 of Wallenia, 13 of Rapanea (i.e. Myrsine), 
i.e. 42 in all. For the Flora of Jamaica I have examined at least 120 Jamaican 
gatherings of Ardisia, 130 of Wallenia, and 50 of Myrsine, i.e. in all more than 300 
gatherings. This increase in material has made evident the distribution of the 
Myrsinaceae in Jamaica, but has not greatly simplified the problems of their classifica- 
tion. Thus, despite much field work, the female flowers of Wallenia clusioides, 
W. crassifolia, W. discolor, W. erythrocarpa and W. xylosteoides and the male flowers 
of W. discolor, W. elliptica, W. erythrocarpa, W. punctulata and W. sylvestris have 
apparently missed collection; they have not been observed in the material studied, 
which, in addition to that of the British Museum (Natural History), London (cited 
below as BM), has included the specimens of the Systematisch-Geobotanisches 
Institut, Universitat Gottingen (GOET), the Institute of Jamaica, Kingston, 
Jamaica (IJ), the Royal Botanic Gardens, Kew (K), the New York Botanical Garden 
(NY) and the University of the West Indies, Mona, Jamaica (UCWI) and the lecto- 
type of Wallenia grisebachii at the Botanische Staatssammlung, Munich. Thanks are 
here expressed to the Directors of these Institutions for their invaluable co-operation 
in lending or making available for study this extensive material. Nevertheless, the 
following account as regards the status and distinctive characters of some taxa is 
necessarily pro visional. Text-figures 21-25 have been drawn by Joanna C. Webb 
(Mrs. D. Erasmus). 



History 

The first members of the Myrsinaceae to be recorded from Jamaica, indeed from 
all America, were Ardisia tinifolia and Wallenia laurifolia gathered by Hans Sloane 
in 1688 or 1689. He collected the Ardisia " on the Mountains near Mr. Elletson's 
Plantation in Liguanee ", i.e. in the parish of St. Andrew, Jamaica, and published 
it in his Catalogus Plantarum: 169 (1696) and Voyage to the Islands Madera etc. 
2 : 98, t. 205 f. 2 (1725) as " Jasminum forte, arboreum, f oliis laurinis obtusis latioribus 



I 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 147 

atrovirentibus, flore pentapetalo racemoso purpureo reflexo ". Concerning this, 
Sloane wrote: " This Tree riseth to about thirty Foot high, having a Clay or Ash- 
colour'd pretty smooth Bark; its Twigs are set about with Leaves which are very 
smooth, of a dark green Colour, having a quarter of an Inch long Footstalks, being 
four Inches long and two broad in the Middle, where broadest, having an eminent 
middle Rib. The Flowers stand on several little Branches, being pentapetalous, 
purplish, bow'd back, having yellow Stamina in their Middle." Olof Swartz found 
the same species " in aridis submontosis Jamaicae " during his stay in the West 
Indies from 1784 to 1786, and gave it the binomial Ardisia tinifolia when founding 
the genus Ardisia in his Nova Genera et Species Plantarum seu Prodromus : 48 (1788). 
Swartz referred five species to Ardisia: 1. A. tinifolia from Jamaica; 2. A. coriacea 
(now regarded as conspecific with A. tinifolia), its provenance vaguely recorded by 
him as " India occidentalis "; 3. A. serrulata {now Parathesis serrulata (Swartz) Mez) 
from Hispaniola, these three being of his own collecting ; 4. A. laterifolia (now Stylo- 
gyne laterifolia (Swartz) Mez), vaguely recorded as from "India occidentalis"; 
5. A. parasitica (now Grammadenia parasitica (Swartz) Griseb.) from " Montserrat " 
(later corrected to Dominica) ; these last two were known to him only from specimens 
in the herbarium of Sir Joseph Banks, now at the British Museum (Natural History), 
as the asterisk against the diagnosis indicates. Later authors have accepted A. 
tinifolia as the type-species of Ardisia. Swartz's description of the species in his 
Flora Indiae Occidentalis i : 468 (1797) is fairly detailed and accompanied by a 
reference to his " Icon. fasc. 3 ". Unfortunately, this third fascicle of his I cones 
Plantarum incognitarum (fasc. i, 1794; fasc. 2, 1800) was never published, although 
Swartz sent the illustrations and text to Schreber in Erlangen for publication. 
Schreber died in 1810, Swartz in 1818; the Palmische Verlagsbuchhandlung in 
Erlangen published a note in Flora (Regensburg) 3 : 144 (March 1820) offering these 
illustrations to whoever would undertake their publication, but no one responded. 
Their subsequent history is unknown until about 1890 when the Berlin botanist 
Ignaz Urban acquired a volume containing presumably all that had survived of 
Swartz's 200 drawings of West Indian plants. They are cited in Urban's Symholae 
Antillanae (cf. i : 164 (1898)) as " Sw. Ic. ined ", and in Fawcett & Rendle' s i^Zom of 
Jamaica (i : xviii, etc. (1910)) as " Sw. Icon, ined: ". They could not be traced 
at the Botanisches Museum, Berlin-Dahlem after the 1939-45 war and hence were 
believed to have been destroyed by bombing in March 1943, as stated in Sitwell and 
others, Great Flower Books : 77 (1956). Luckily, however. Urban had sold this 
precious volume in 1922 (cf. Journ. Bot. (Lond.) 60 : 361 (1922)) to C. A. Lindman 
in Stockholm and it is now preserved in the library of the Royal Swedish Academy 
of Sciences; Plates 6 and 7 are reproduced here from these drawings by permission 
of the Royal Swedish Academy of Sciences, to which grateful acknowledgement is 
made. Sets of photographs of Swartz's unpublished drawings have been deposited 
in the libraries of the British Museum (Natural History), London, the Institute of 
Jamaica, Kingston, Jamaica, and the Hunt Botanical Library, Pittsburgh, U.S.A. 
Swartz's coloured drawings representing Ardisia tinifolia and A. coriacea, together 
with A . serrulata and A . parasitica, leave no doubt as to the application of the names. 
Swartz applied the name A. coriacea to a Jamaican plant (without evident veining 



148 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

of the leaves but otherwise very similar to his A. tinifolia) which was described by 
Mez in 1901 as A. harrisiana; hence, as pointed out by Urban (Symb. Antill. 3 : 330 
(1902); op. cit. 8 : 519 (1921)), Mez used the name A. coriacea wrongly for a species 
of the Lesser Antilles, i.e. A. ohovata [A. guadalupensis) . 

According to Sloane's account (Voy. Nat. Hist. Jam. i : 234 (1707)) of the 
Jamaican plant now known as Wallenia laurifoUa, but recorded by him as " Bryonia 
nigra fruticosa, foliis laurinis, floribus, racemosis, speciosis ", he " gathered it, if 
I rightly remember, in St. Maries, near Cabe9a del oro in the North-side of this Island 
amongst the Woods " ; he described the flowers simply as " at the top of the Branch . . . 
standing in a bunch together, being many very beautiful and small ". On this 
species Swartz in 1788 (Nov. Gen. & Sp. PI. : 2, 31) founded the genus Wallenia, 
naming it in honour of Matthew Wallen, from whom Swartz received much hospitality 
while in Jamaica. Meanwhile, Jacquin had published in his Seledarum Stirpium 
Americanarum Historia : 17 (1763) the provisional desigmXion Petesioides laurifolium 
referring to incomplete material collected by him in Santo Domingo (Haiti) between 
1755 and 1759, and Petesioides was accordingly cited in the 1952, 1956, and 1961 
editions of the International Code of Botanical Nomenclature, under No. 6304, as 
a nomen genericum rejiciendum necessitating the conservation of SwdiViz's Wallenia. 
Linnaeus had, however, ruled in Critica hotanica : 34 (1737) that " nomina generica 
in aides desinentia, e foro Botanico releganda sunt " (generic names ending in aides 
are to be banished from the botanical forum). Hence, when Linnaeus himself used 
designations such as Bacchariaides, Bannister aides, Carnutiaides, Euanymaides, 
Hibiscaides, Oxycoccoides, Senecioides etc. in his Flara Zeylanica (1747) for obscure 
plants and when loyal followers of Linnaean method and nomenclature such as 
Jacquin, Loefling and Rottboll used designations such as Celasioides, Diadiades, 
Maipighiaides, Sideraxyloides , Staehelinoides and Viscaides, they did not intend them 
to be permanent names for the genera concerned but merely as provisional or token 
names indicating resemblance; they intended that these -aides designations should 
later be replaced by correctly formed names. Jacquin's description of his Petesiaides 
indicates that he possessed only a specimen with functionally female flowers; he 
believed that it represented a new genus which could not be properly established 
until fruits became available; accordingly, in his own words, " ego interea ab habitu 
nomen fluxum imposui " (I meanwhile, from its habit of growth, have given it a 
perishable name). Such token designations are in a different category from the 
pre-Linnaean generic names ending in -aides, such as Capnaides, Ficaides, Nymphaides, 
brought into post-Linnaean use as definite generic names by Miller, Hill, Adanson, 
Medicus and others. In a like spirit Rottboll in 1772 described Schaenaides and 
Scirpaides as new genera to which names would be given later ; in 1773 he named the 
first Kyllinga, the second Fuirena. Swartz thus acted rightly according to the 
procedure of his times in replacing Petesiaides by Wallenia and Jacquin evidently 
approved Swartz's action, for he himself adopted the name Wallenia in his Plantarum 
rariarum Harti Caesarei Schoenbrunnensis Descriptianes i : 13, t. 30 (1797). Valid 
publication of " Petesiades Jacq." as a real generic name dates from Otto Kuntze's 
adoption of it in 1891 (Revis. Gen. PI. 2 : 402). 

In 1788 Swartz named, as Samara cariacea (Nov. Gen. & Sp. PI. : 32), a third 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 149 

Jamaican member of the family, referring it to the Linnaean genus Samara, which is 
now included in Emhelia Burman f. This species, found " in sylvis montium 
Jamaicae australis ", is congeneric with Rapanea guianensis Aublet, as Swartz noted, 
but Rapanea itself is not well separated from Myrsine L. 

The collections including Myrsinaceae made in Jamaica during the first half of the 
nineteenth century by Macfadyen, Alexander, McNab, Nathaniel Wilson, March, and 
Purdie and used by Grisebach in preparing his Flora of the British West Indian 
Islands : 392-397 (1861) provided the material for publication there of Wallenia 
clusiifolia, W. venosa, Ardisia xylosteoides, and A. clusioides, a.nd the record of A. 
humilis (actually A. solanacea) as naturalized in Jamaica. Grisebach's book (cf. 
Stearn, 1965) seems to have given the impression for many years that no further 
collecting need be done in the West Indies. WiUiam Harris's collecting from 1894 
onwards brought about a new era in the botanical exploration of Jamaica, revealing 
the existence there of a multitude of new species which had escaped the notice of 
earlier collectors (cf. Urban, Symb. Antill. 6 : 70-86 (1909)). In 1895 Krug & Urban, 
working out Harris's collections from the Blue Mountains, Jamaica, published two 
further species, Myrsine acrantha and Ardisia densiflora. Mez in 1901 recorded from 
Jamaica four species of Ardisia, A. harrisiana being proposed as new, eight species 
of Wallenia, and four of Rapanea. At the same time he defined more satisfactorily 
the genera, by transferring to Wallenia some species which Grisebach had included 
in Ardisia, although he separated Rapanea from Myrsine, reducing the latter to four 
African and Asiatic species typified by M. africana L. 

The acquisition of new Jamaican material, mostly collected by Harris and by N. L. 
Britton of the New York Botanical Garden, and the re-study of the older material 
seen by Mez led Urban between 1908 and 1915 to describe as new Ardisia dictyoneura, 
A. troyana, Wallenia calyptrata, W. corymbosa, W. discolor, W. elliptica, W. erythro- 
carpa, W. punctulata and W. sylvestris and Britton to describe Petesioides sub- 
verticillata (now Wallenia subverticillata) . Urban's descriptions are wonderfully 
accurate, clear and detailed and are accompanied by helpful critical notes in which 
he made the most of inadequate material, but as usual he never provided a key to 
aid their recognition. The construction of such a key has indeed proved very 
difficult, partly because some taxa are only known in one of their three possible 
states — thus of the 16 species of Wallenia distinguished here, only six have been 
collected in their staminate, pistillate and fruiting states — but also because of the 
variation among these plants, which makes every gathering seem a little different 
from other gatherings. This is discussed further below under Wallenia. 

Geographical Distribution 
Four, if not five, of the Jamaican members of the Myrsinaceae occur outside the 
island. These are Ardisia compressa, which extends from Mexico to northern South 
America, Myrsine coriacea, which extends from Cuba over Jamaica, Hispaniola, 
Puerto Rico and the Lesser Antilles, and from Mexico over Central America, to 
Bolivia and Argentina, Wallenia laurifolia, which occurs also in Cuba and Hispaniola, 
and W. subverticillata, which also occurs in Cuba. Ardisia densiflora is closely allied 



I50 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

to A. spicigera of Mexico and A. gentlei of British Honduras; indeed, according to 
Lundell (1966a) the three are conspecific. The ornamental Asiatic species A. 
solanacea (Text-fig. 8) has become naturalized in several places. All the others appear 
to be endemic to Jamaica, although Myrsine acrantha may probably be found on other 
islands; specimens of it are likely to be referred to Myrsine {Rapanea) guianensis. 
Members of the family may be found in Jamaica at any altitude from sea-level to the 
top of Blue Mountain Peak (7,400 ft., 2,250 m.). 

Although it would be premature to define types of distribution within Jamaica on 
the basis of present material, this is adequate to indicate relative ranges. Thus 
Myrsine acrantha (Text-fig. 20), M. coriacea (Text-fig. 19), Ardisia tinifolia (Text- 
figs. 1-4), Wallenia laurifolia (Text-fig. 17), and W. venosa (Text-fig. 14) have 
comparatively wide ranges extending both east and west. A few are localized in the 
east, i.e. Ardisia hrittonii (Text-fig. 5), Wallenia sylvestris (Text-fig, 18), and W. 
suhverticillata (Text-fig. 9). Some others are restricted to the higher parts of the 
Blue Mountains, i.e. Ardisia densiflora (Text-fig. 6), Wallenia calyptrata (Text-fig. 13), 
W. crassifolia (Text-fig. 10), and W. fawcettii (Text-fig. 16). As far as is known, 
Ardisia byrsonimae (Text-fig. 5), ^. urbanii (Text-fig. 5), Wallenia corymbosa (Text- 
fig. 11), W. elliptica (Text-fig. 9), and W. punctulata (Text-fig. 13) are restricted to the 
middle of the island. Ardisia dictyoneura (Text-fig. 6), Wallenia grisebachii (Text- 
fig. 15), and W. purdieana (Text-fig. 16) occur both in the middle and west, Ardisia 
compressa (Text-fig. 7), Wallenia clusioides (Text-fig. 9), and W. erythrocarpa (Text- 
fig, 12) only in the west. As there has obviously been no bias in making the collec- 
tions here analysed, the species such as Ardisia brittonii (Text-fig. ^), A. byrsonimae 
(Text-fig. 5), ^. urbanii (Text-fig. 5), Wallenia discolor, W. elliptica (Text-fig. 9), 
W. erythrocarpa (Text-fig. 12), and W. punctulata (Text-fig. 13), which are known 
only from single gatherings or from gatherings made at the same place, must be of 
very limited range. 

Specimens are cited below under Jamaican parishes, proceeding usually from 
west to east or from north to south within the parish. 



Family Characters 

The family Myrsinaceae as represented in the West Indies consists of small trees 
or shrubs, with the leaves usually alternate, rarely whorled, and without stipules. 
The flowers are regular, bisexual or unisexual, with parts in fours or fives, and are 
grouped in axillary or terminal racemes or panicles sometimes condensed into 
fascicles or umbels. The calyx is persistent, with segments free or basally connate, 
imbricate or valvate. The corolla is usually gamopetalous but in Myrsine acrantha 
parted to the base. The stamens are inserted on the corolla, opposite to and the 
same number as the lobes of the corolla, with the anthers opening introrsely by 
lengthwise slits or apical pores. The ovary, superior in West Indian genera, is 
one-chambered with a free central placenta, into which the few or many ovules are 
sunk, appearing as slight or distinct bulges on its surface ; the style may be short or 
long, with the stigma undivided or lobed or fimbriate. The fruit in West Indian 
genera is a more or less globose drupe. The family as a whole includes about 900 



< 


^ 


-A 


u 


f . 


7 


>- 


1 , 




^ 


< ' 1 y. 


2 


/i u ) 


< 


I ' »j 




/J-*^ 




ft 


;^ 


1 


;:£ 





1 




tif 










^ 1 



«>: 



^ 
§ 



< 

< 

< 


( 


"] 




( 




i»--''\ i f' 


i 




/ 




1 '• r 





t 



TS 



to 



"8, "«>: 


< 

U 


[ .' 


•^ 


2 
< 


J... [. /0 
1 '■■•■•'• -■'' • 


/ 






• y 


( 


i / ^ "•..- 


...... 


7'-''"'"'b^ 




[•I* 





5S 



J AM A I C A 


;< 


"^ 


/)£ 


i 


I 


i (• ..-■ 


^--V>i^ 


) 


— Y 




• ••!«....--;■■ 


3 




txT 


: 





ft 



«^ 



"a, ■'; 


>■ 




< 
U 

<* 


/; 1 \ 


2 

< 




I 




] J (Mi 


;|' 


i 


^ /• : *■■■ 

■ rf^'i 


y 


rmv-i y 




\S7 







^ 






< MirA 



?>: 



< 

u 
■ < - 




^ 


2 

< 


—f : ,5' "■•■' 


/ 




1 i 'l^: ,,. 


T 


( 


:•).:..;.---■'■■ 


^ . 


^;;:j,;«:^ 




M 





"»>: 







< 

u 

<• 


^ 


^ 


2 
< 




/ 




/ * ^ 


T 


( 


/ ^ 


/ 




3^y 


y 






w 


^r 





s 



< 

• < - 


f ■' 


^ 


2 
< 




< 


( 


1 , - ' 'T ^V^ 


^ 




c •!; 








"Si 


°« 


i:' 






< 

U 


f 




■ \ 




<• 


/ 


^ 






2 


J 


.■' 






< 


J 


/ 


1 






/) 


-.../ 




\ 




i''^.\;. 


.:...-•■ 


.1 


) 


I 


; j--; 


---'■■ 




•> 




i 


■--—y 




J 




:5-^.. 


J^ 




) '■ ' T' 






Sp 




(••; ; 











"■S 



•TS 



< 

u 
< - 


> '' 


< 


s • ' 


I 






{ /•■; 

/-.i.» 


:^ 


) 


i 

V 


f :.f -:..::: 


y 




w^ 









•s. ■«^- 








< 




^ \ 








. u 












~ 


j ,' 








<*» 


■ < ■ 






I, 




\ . ■' 




) 




2 


e ''■ I 










< 


/ : *• 


' / 








" 


r^ ? • 












V ' '■' 




) 


r^ 
r^ 




[.— ^•'' ■> 


(^ 


/ 






] : \.j _ 


^^..V 


M-5^ 








> !.-*•••■' 




y' 








/ • V 




j/ 




"-i 




t. .■ ■; J.,.--' ' 




( 


•-) 




i • 




) 






: •\....,..--"' 










1 


:..../ i • 


-y 


\ 




00 


V 


I..'-';' '^"T^ 






00 

1^ 




1 -' r^ 




S| 






I*' 




^ 










\^.; 










^p 










o 


-. 




■«^ 




•-. 



- s 



< 

<<• 


f / 


'\ 


2 
< 




i 




/%■ 


f 


V 




y 









JS^ 



»>; 



< 




^ 


u 


f / 


] 


. <; . 


< ^ 


\ 


2 
< 




^ 




I'Br-- 


-^ 




/ • ';>-'• 


'' ( 


( 




3 




J •-'' r 


!f 




/ ! : 








w 







"»i; 



s 



I 

I: 



s> ■«>; 1 


< 










Ci 






2 


f\ L 




\ 


< 


M. 




} •- 


r*. 




S 


f 




/. i...Si-..-i" 


... 


■° 


"V 


; •/' 1 • - 


— 


1 


«o 


"I*--]' " : '>^ 


• 



w 


r-' 


[I r 




r^ 




w 




f^- 


"2, '»>; 


-v 



Sv '?s 


>• 






< 




^ 




L> 


(^. 


• 1 




_ 


I ' 






<f 


f '. 






2 


1 ^V 


} 


< 




^ 


/J~' 


^ 


] 




fl> 


^-' 


! 


I 


;£ 






1 '' J 




•-t 




( • • 




- 






V: / 




I 






^■^vy 




« 










'^ 





>s 




< 

• < - 


/-^ 


V 


2 

< 


^^ ';-■"■'■?■' 


^^ 


> 




/......'■ \ • 

1 • • ;j 

/ '-* .■'■" 




r - 


( 








y 






( •!; 


1 



s 



< 

u 
■ < - 

< 






r 




/ :• re.:---' 


:^ 


/ 


/ : • 

;.•;:;:-::■;;''' 


.... 




/ 


( 




^ 




22 





>^ 



156 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

Species and occurs in the tropics and sub-tropics of the Old and the New World. 
Of its 30 or so genera, nine are represented in the West Indies as a whole but only 
three in Jamaica. Those absent from Jamaica are Parathesis (Cuba, Hispaniola, 
Puerto Rico), Stylogyne (Lesser Antilles), Conomorpha (Lesser Antilles), Weigeltia 
(Lesser Antilles), Cyhianthus (Trinidad) and Grammadenia (Puerto Rico, Lesser 
Antilles). 

Key to Jamaican Genera 

Flowers in usually terminal (sometimes axillary and then long-pedunculate) racemes 

or panicles; style slender; stigma inconspicuous. 

Flowers bisexual, with stamens and style both well developed in the same flower; 

corolla c. 6-12 mm. long; ovules numerous, scattered in several series over 

the placenta . . . . . . . . . i. Ardisia 

Flowers unisexual, dioecious, the male flowers with well developed exserted 
stamens, the female flowers with abortive included stamens but exserted 
styles; corolla 1-5-6 mm. long; ovules 3-4 in a single horizontal row on the 
placenta . . . . . . . . . .2. Wallenia 

Flowers in umbellate clusters on very short, scaly axillary spurs ; style short or none 
(in American species) ; stigma conspicuous . . . -3. Myrsine 

I. ARDISIA Swartz 

Ardisia Swartz, Nov. Gen. & Sp. PI. : 3, 48 (1788) ; Fl. Ind. Occ. i : 467 (1797) ; 

nomen conservandum No. 6285. 
Lectotype: A. tinifolia Swartz. 

Key to Jamaican Species 

Inflorescences all axillary; pedicels 1-4 cm. long, clustered almost umbellately at 

the end of the peduncle; anthers 5-6 mm. long . . . 8. ^. solanacea 

Inflorescences terminal (sometimes with a few axillary inflorescences below) ; pedicels 

mostly 1-7 mm. long, sometimes to 12 mm. long but then racemosely arranged 

along the lateral peduncle; anthers 1-5-5 ni^- long: 

Lateral peduncles compound, i.e. bearing 3 to 4 secondary peduncles each 

umbellately 2-6-flowered; leaves 7-16 cm. long, the blades narrowly obovate 

to oblanceolate ; filaments of stamens 0-5 mm. long . . y. A. compressa 

Lateral peduncles simple, i.e. directly bearing 3 to 28 pedicelled flowers but no 

secondary peduncles; leaves mostly less than 12 cm. long, the blades various; 

filaments of stamens 1-5 mm. long: \ 

Leaves closely clustered at the tips of shoots, less than 5-5 cm. long; inflorescence 

contracted, less than 4 cm. long and broad, the branches 3-6-flowered, the 

pedicels 1-4 mm. long; central Jamaica . . . 6. A. byrsonimae 

Leaves more widely spaced, some or all more than 5-5 cm. long; inflorescence 

more than 4 cm. long and broad, the branches usually many-flowered, the 

pedicels 2-15 mm. long: 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 157 

Inflorescence somewhat fastigiate, the densely many-flowered branches 
ascending at c. 40°; corolla c. 5 mm. long; anthers c. 1-5 mm. long, much 
shorter than the filaments; Blue Mountains . . 5. A. densiflora 

Inflorescence looser, the branches spreading at c. 50-60° ; corolla 7-12 mm. 
long; anthers 2-5-5 mrn. long, equalling or longer than the filaments: 
Pedicels 1-4 mm. (rarely to 5 mm.) long; leaf -blades with well-marked 
reticulate venation . . . . . 4. ^. dictyoneura 

Pedicels 4-15 mm. long; leaf-blades not markedly reticulate: 
Sepals 2-5-4 mm. broad: 
Anthers 4-5-5 mm. long, apically dehiscent; sepals almost or quite 
glabrous; eastern Jamaica . . . . 2. A. hrittonii 

Anthers 3-3-5 mm. long, laterally dehiscent almost to the base; sepals 
ciliate; western Jamaica. .... ^.A.urbanii 

Sepals 1-2 mm. broad: 

Sepals ovate or broadly ovate : 

Anthers 3-3-5 mm. long . . . . 1. A. tinifolia var. i 

Anthers 4-5 mm. long .... A. tinifolia var. 2 

Sepals mostly narrowly ovate or oblong-lanceolate : 

Leaf -blades mostly less than 6 cm. long; sepals 2-2-5 "i™- long 

A . tinifolia var. 3 
Leaf-blades 7-12 cm. long; sepals 2-5-3 rni^- lo^^g A. tinifolia var. 4 

Subgenus i. ARDISIA 
Ardisia subgenus Ardisia; subgenus typicum generis. 

Ardisia Swartz, Nov. Gen. & Sp. PI. : 3, 48 (1788), sensu stricto. 
Pickeringia Nutt. in Journ. Acad. Sci. Philad. 7 : 95 (1834). 
Ardisia sect. Euardisia Endl., Gen. PI. : 736 (1839). 

Ardisia subgenus PicAmw^m (Nutt.) Mez in Urban, Symb. Antill. 2 : 396 (1901); in Engler, 
Pflanzenr. IV. 236 : 79 (1902). 

Lectotype: A. tinifolia Swartz. 

Inflorescence always terminal; branches simply racemose. Stamens with con- 
spicuous filaments. Species all American. 

I. Ardisia tinifolia Swartz, Nov. Gen. & Sp. PI. : 48 (1788); Fl. Ind. Occ. i : 468 
(1797). — Griseb., Fl. Brit. W. Ind. Is. : 396 (1861). — Mez in Urban, Symb. 
Antill. 2 : 400 (1901); in Engler, Pflanzenr. IV. 236 : 83, fig. 11 F-K (1902). 
(Text-fig. 23G, H; pi. 6, 7.) 

Ardisia coriacea Swartz, Nov. Gen. & Sp. PI. : 48 (1788); Fl. Ind. Occ. i : 470 (1797). 
Ardisia harrisiana Mez in Urban, torn. cit. : 401 (1901); in Engler, torn. cit. : 83 (1902). 
Ardisia troy ana Urban, Symb. Antill. 5 : 455 (1908). 
" Jasminum forte, arboreum, foliis laurinis obtusis latioribus atrovirentibus, flore pentapetalo 

racemoso purpureo reflexo " Sloane, Cat. PI. Jam.: 169 (1696); Voy, Jam. Nat. Hist. 2: 

98, t. 205 fig. 2 (1725). 

The species Ardisia tinifolia as here accepted includes plants diverging in form and 
size of leaf, in number of flowers, shape of sepals and length of anthers. Selected 
specimens, e.g. Harris 8^41 (type-collection of A. troy and) from near Troy, Central 



158 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

Jamaica, with very small leaves, Proctor 16621 from Orange River Valley, St. James, 
with comparatively elongated sepals, and Loveless iigS from the Long Mountain, 
near Kingston, with long anthers, could reasonably be regarded as exemplifying 
independent species and were indeed treated by the writer as such until the difficulty 
of finding stable combinations of characters and of allocating specimens intermediate 
in one character or another to these supposed taxa made it more practical to place 
them all in one species, within which four major populations (Text-figs. 1-4) are 
more or less distinguishable though not easily defined. They are simply given 
numbers below. Swartz's original diagnosis of his A. tinifolia in 1788 is brief but 
the more detailed description in his Flora Indiae Orientalis, and his unpublished 
illustration (PI. 6, see p. 147) and authentic specimens (BM) clearly establish the 
identity of the plant he had in mind. A representative modern gathering is Harris 
i2o8y, collected between Constant Spring and Bardowie, St. Andrew, with veining 
on the lower surface of the leaf plainly evident. It typifies the population (var. 
I below) found in the southern parishes of Jamaica, with an altitudinal range of 
50 ft. (15 m.) to 3,500 ft. (1,070 m.). Here belongs Swartz's A. coriacea with veining 
on the lower leaf-surface faint or not evident, likewise represented by an unpublished 
drawing (PI. 7). Diverging from typical A . tinifolia principally by having somewhat 
longer anthers is the population (var. 2 below) characteristic of the Long Mountain 
near Kingston, at 500 ft. (150 m.) to 1,000 ft. (300 m.). Small short-petioled leaves 
evidently veined characterize plants of central Jamaica (var. 3 below) named A. 
troy ana by Urban, which also differ from typical A . tinifolia in having shorter branches 
of the inflorescence. These occur at 700 ft. (210 m.) to 2,500 ft. (750 m.). Their 
range slightly overlaps that of a population (var. 4 below) with larger longer-petioled 
leaves, looser many-flowered inflorescences and slightly shorter filaments, found in 
the northern parishes of Jamaica from almost sea-level to 700 ft. (210 m.). They 
occur in the same area at Sherwood Content, Trelawny, and merit particular study 
there in the field. 

Together the plants here placed in Ardisia tinifolia thus give the impression of 
a population within which morphologically and geographically differentiated taxa 
have been formed but have not achieved complete independence. 

Var. 1. 

A. tinifolia Swartz, loc. cit. (1788). 
A. coriacea Swaxtz, loc. cit. (1788). 
A. harrisiana Mez, loc. cit. (1901). 

Leaf-blades mostly elliptic, sometimes narrowly obovate, the apex obtuse or 
acute, 5-12 cm. long, 1-5-7 cm. broad. Sepals imbricate, ovate or broadly ovate, 
rounded, c. I-5-2-5 mm. long, 1-2-2 mm. broad. Anthers 3-3-5 mm. long. 

St. Elizabeth: Proctor 15419 (BM; IJ), Proctor 11344 (BM; IJ). 

Clarendon: Proctor 22^39 (BM; IJ), Proctor g';i2 (IJ), Proctor 10233 (IJ), Steam 
1044 (BM), Adams 8424 (UCWI). 

St. Ann: Alexander (K). 

St. Catherine: Proctor 9310 (BM; IJ), Proctor 8373 (IJ), Proctor 22y3g (IJ). 

St. Andrew: Britton 3461 (NY), Steam 5g (BM), Proctor 11428 (IJ), Harris sgyo 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 159 

(BM; NY), Harris 11 137 (NY; UCWI), Newill 5449 (IJ), Harris 5926 (NY), Harris 
12081 (BM; NY; UCWI), Harris 9033 (BM; NY; UCWI), Powell 850 (IJ), Proctor 
23694 (IJ), Harris 6600 (BM; NY; UCWI : type-collection of Ardisia harrisiana Mez), 
Britten 2987 (NY), Proctor 2451 1 (BM), Proctor 23982 (BM). 
St. Thomas: Adams 12579 (UCWI). 

Var. 2. 

Leaf-blades elliptic or broadly elliptic, a few broadly obovate, the apex obtuse or 
rounded, 4-9 cm. long, 2-6 cm. broad. Sepals imbricate, ovate or broadly ovate, 
c. 2-2-5 rnni- loiig' i"5-2 mm. broad. Anthers 4-5 mm. long. 

Kingston and St. Andrew: Britton 377 (NY), Robbins 1496 (UCWI), Proctor 7386 
(IJ), Webster 4986 (BM), Loveless 1198 (UCWI), Yuncker 17315 (BM), Anderson & 
others 626 (UCWI). 

Var. 3. 

Ardisia troy ana Urban, loc. cit. (1908). 

Leaf-blades elliptic, the apex obtuse, c. 3-10 cm. long, 1-5-6 cm. broad (usually 
less than 6 cm. long, 3-5 cm. broad). Sepals narrowly ovate, 2-2-5 mm. lo^ig. 
I- 1-5 mm. broad. Anthers 3-3-5 mm. long. 

Trelawny: Harris 8741 (BM; NY; UCWI: type-collection of A. troy ana Urban), 
Britton 605 (NY), Harris 9473 (NY; UCWI), Howard, Proctor & Steam 14677 (BM), 
Howard, Proctor & Steam 14679 (BM; IJ), Proctor 11060 (BM; IJ), Proctor 22543 
(BM; IJ), Proctor 24458 (BM; IJ), Proctor 25676 (IJ). 

Clarendon: Harris 12769 (NY; UCWI), Proctor 9772 (IJ), Proctor 10233 (IJ). 

Manchester: Proctor 25625 (BM; IJ). 

Var. 4. 

Leaf -blades narrowly obovate or elliptic, the apex acute or shortly acuminate, 
5-12 cm. long, 2-6 cm. broad. Sepals scarcely or not imbricate, ovate to obovate- 
lanceolate, obtuse, c. 1-5-2-5 mm. long, 1-1-5 rn^^- broad. Anthers c. 3 mm. long. 

Hanover : Proctor 11280 (BM ; I J) . 

Westmoreland: Harris 10220 (BM; K; NY; UCWI). 

St. James: Harris 10318 (BM; NY; UCWI), Britton & Hollick 2355 (NY) , Proctor 
16621 (BM; IJ), Proctor 24292 (BM; IJ). 

Trelawny: Proctor 11062 (BM; IJ). 

St. Mary: Loveless 2437 (UCWI), Proctor 7546 (IJ), Yuncker 17838 (BM). 

2. Ardisia brittonii Stearn, sp. nov. (Text-fig. 21; PI. 8a) 

Arbor parva ad 4-5 mm. alta, raww/^shornotinisprimo ferrugineispulverulentisque, 
annotinis nigrescentibus glabrisque. Folia dissita, c. 4-8 mm. inter se distantia, 
magna, patentia, longipetiolata ; lamina oblongo-elliptica, apice obtusa, margine 
Integra, basi cuneata in petiolum attenuata, 8-17 cm. longa, 3-8 cm. lata, glabra, 
utrinque punctata, coriacea, venis numerosis utrinque prominulis marginem versus 
reticulate anastomosantibus ; petiolus 1-2-2 cm. longus. Inflorescentiae terminales 
multiflorae laxae, c. 15 cm. longae, 10 cm. latae, supra glabrae, interne minute 

BOT. 4, 4 9 



i6o 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 



pubescentes, ramulis patentibus ad 9 cm. longis racemose 8-12-floris; pedicelli 
inferiores ad 1-5 cm. longi. Calyx c. 4-5-5 mm. longus; sepala imbricata parum 
inaequalia, ovata vel late oblonga, rotundata, c. 4-4-5 mm. longa, 2-5-3 i^m- lata, 
punctata, margine glabra. Corolla rosea, c. 1-3 cm. longa; petala in 1/4 altitudinis 
coalita, apice obtusa, parte libera oblonga c. 4 mm. lata, per anthesin revoluta, 
nigro-punctata et lineolata. Stamina quam petala breviora, c. 1-5 mm. supra 
coroUae basin inserta; antherae anguste triangulares, c. 4-5-5 mm. longae, apice 
obtusa rimis brevibus dehiscentes, basi cordatae ; j^/amew/a 3 mm. longa. Ovarium 
late ovoideum; stylus c. 9 mm. longus, stigmate minuto. 




Fig. 21. Dissected flower of Ardisia brittonii Stearn {Harris 6- Britton 10558), X 3. 

St. Thomas : Bath to Cuna Pass, Sept., 1908, Harris & Britton 10558 (NY, holotype: 
BM; K;UCWI). 

This species of eastern Jamaica has been named in honour of Nathaniel Lord 
Britton (1859-1934), founder of the New York Botanical Garden and author of 
many botanical works on the eastern United States and the West Indies; see 
National Acad. Sci. U.S.A. Biogr. Mem. 19 : 147-202 (1938) for a biography by 
E. D. Merrill and a bibliography by J. H. Barnhart. 

Ardisia brittonii differs from the widespread A. tinifolia principally in its broader 
sepals and in having somewhat larger flowers and longer anthers than is usual in 
A. tinifolia. It agrees more closely with A. urbanii of central Jamaica but may be 
distinguished from that by its glabrous calyx and its longer anthers, which dehisce 
only at the apices, 

3. Ardisia urbanii Stearn, nom. nov. (Text-fig. 22.) 

Ardisia rosea Urban, Symb. Antill. 5 : 456 (1908); non A. rosea King & Gamble (1905). 

Trelawny: Harris 9419 (BM; NY; type-collection), Harris 9418 (BM; NY; 
UCWI), Britton 659 (NY). 

Ardisia urbanii is known only from collections made at Troy on the southern side 
of the Cockpit Country by Harris & Britton in 1906 at about 1,600 ft. (530 m.). The 



i 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 



i6i 



size of the calyx (with sepals 3-5 mm. long, 2-5-4 ^^- broad) distinguishes it from 
A. tinifolia; the ciliated sepals and the anthers dehiscent to the base distinguish it 
from A. hrittonii of eastern Jamaica. According to Harris, it forms a small tree to 
25 ft. (7-5 m.) high, with rosy flowers. It is named in honour of Ignaz Urban (1848- 
1936), the author of many important works on the West Indian flora; see Ber. 
Deutsch. Bot. Ges. 48 : (205)-(225) (1931) for a biography and bibliography by 
Th. Loesener. 



V 




Fig. 22. Dissected flower of Ardisia urbanii Steam (Britton d^g), x 4. 



4. Ardisia dictyoneura Urban, Symb. Antill. 6 : 28 (1909). 

Hanover: Harris 10345 (BM; NY; UCWI; type-collection), Proctor 11316 (BM; 
IJ), Britton 2277 (NY). 

Westmoreland: Proctor ^321 (BM; IJ). 

St. James : Proctor 22gyg (BM ; I J) . 

Trelawny: Proctor 10610 (BM; IJ). 

Manchester: Proctor 11028 (BM; IJ), Proctor 11036 (BM; IJ), Proctor i82gg (IJ), 
Steam 351 (BM). 

St. A-n^n : Purdie (K), Proctor 22835 (BM; IJ), Proctor 26724 (BM), Steam 567 (BM), 
Proctor ii8g6 (IJ), Howard & Proctor i503g (IJ), Proctor J4g4 (IJ). 

St. Catherine: Howard & Proctor 13620 (IJ; NY). 

Ardisia dictyoneura is a species of western and central Jamaica at 1,000-3,000 ft. 
(300-900 m.). The type collection was made by Harris in 1908 at Fray Woods in 
Hanover. It has often been collected in fruit, only twice in flower, and is notable 
for its profusely and finely reticulate-veined coriaceous mostly obovate or narrowly 
obovate leaves and its comparatively large fruits, 6-5-8 mm. in diameter when 
mature, borne on pedicels only 1-4 mm. long. 

BOT. 4, 4 9§ 



i62 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

5. Ardisia densiflora Krug & Urban in Notizbl, Bot. Gart. & Mus. Berlin i : 79 
(1895). — Urban, Symb. Antill. i : 385 (1899); op. cit. 6 : 28 (1909). — Mez in 
Urban, Symb. Antill. 2 : 398 (1901); in Engler, Pflanzenr. IV. 236 : 82 (1902). — 
C. L. Lundell in Fieldiana, Bot. 24 (8) : 141 (1966). (Text-fig. 23F.) 

Ardisia spicigera Donn. Smith in Bot. Gaz. 27 : 434 (1899), fide Lundell, loc. cit. (1966). 
Ardisia gentlei C. L. Lundell in Field & Lab. 13 : 11 (1945), fide Lundell, loc. cit. (1966). 

St. Andrew: Harris ^22 j (K; UCWI), Harris 5431 (BM; K; type collection), 
Harris 7657 (BM ; K ; NY ; UCWI), Harris 6578 (BM ; UCWI), Harris ioo2g (BM ; NY ; 
UCWI), Harris 6077 (NY; UCWI), D. Watt 8225 (NY; UCWI), Maxon & Killip 1343 
(NY), Proctor 2451 1 (BM; IJ), Proctor 25610 (BM; IJ). 

The discovery in 1894 of Ardisia densiflora in the Blue Mountains of Jamaica, on 
which it grows at about 3,000-4,000 ft. (900-1,200 m.), was one of the early results 
of the renewed collecting in Jamaica at the end of the nineteenth century. It is a 
well-marked species with a densely many-flowered inflorescence, the branches of 
which are ascending, almost fastigiate, and which bear up to 25 small white flowers 
on pedicels 1-6 mm. long. The filaments of the stamens are much longer than the 
longitudinally dehiscent anthers. According to Lundell, A. spicigera from Mexico 
and A. gentlei from British Honduras are both conspecific with A. densiflora, which 
thus has a fairly wide but sparse Central American distribution comparable with 
that of Achimenes erecta (Lam.) H. P. Fuchs, a gesneriad restricted in Jamaica to the 
Blue Mountains. 



6. Ardisia byrsonimae Stearn, sp. nov, (Text-fig. 23A-E; PI. 8b). 

Arbor parva ad 4-5 m. alta, ramulis hornotinis ferrugineis pulverulentisque, 
annotinis cinereis glabrisque. Folia pro genere parva, ad apices ramulorum conferta, 
1-5 mm. inter se distantia, ascendentia, brevipetiolata ; lamina obovata, apice 
rotundata interdum emarginata, margine Integra, basi cuneata, c. 3-5-5 cm. longa, 
i*3~3'5 cm. lata, glabra, utrinque punctata, coriacea, venis vix prominulis; petiolus 
2-5 mm. longus. Inflorescentiae terminales fere sessiles multiflorae densae, c. 3 cm. 
longae et latae, primo pulverulentae, ramulis ascendentibus ad 3 cm. longis racemose 
3-6-floris; pedicelli 1-4 mm. longi. Calyx c. 3 mm. longus; sepala imbricata, ovata, 
rotundata, c. 2-2-5 t^t^- longa, 1-5-2 mm. lata, margine scariosa ciliataque. Corolla 
rosea, c. 7-8 mm. longa; petala in 1/4 altitudinis coalita, apice acuta, parte libera 
lanceolata, 2-2-5 mm. lata, per anthesin recurvata, punctata. Stamina quam petala 
breviora, c. 1-5-2 mm. supra corollae basin inserta; antherae anguste triangulares, 
c. 3-5-5 mm. longae, apice mucronatae et rimis brevibus quasi poratim dehiscentes, 
basi cordatae; filamenta c. 2 mm. longa. Ovarium late ovoideum; stylus c. 8-9 mm. 
longus, stigmate minuto. Fructus transverse late ellipsoideus, c. 4-5 mm. altus, 
5-6 mm. latus, glanduloso-punctatus, styli basi c. i mm. longa coronatus. 

Clarendon: Peckham Woods, 760 m., Dec. 1917, Harris i27gg (BM; NY; UCWI), 
same locality, Aug. 1954, Webster & Proctor 5428 (BM), same locality, Jan. 1955, 
Proctor 9743 (BM; IJ), same locality, Dec. 1955, Stearn 9 (BM, holotype; K), Proctor 
1 1408 (IJ). 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 



163 



JCW 




Fig. 23. Ardisia byrsonimae Stearn: A, fruiting specimen [Steam 9), x i; B, buds {Proctor 
9743), X 2; C, dissected flower [Stearn g), x 3; D, anther [Stearn g), x 5; E, fruit 
[Stearn g), x 4. Ardisia densiflora Krug & Urban: F, dissected flower [Harris I002g), 
X 4. Ardisia tinifolia Swartz; G, dissected flower [Harris 12081), x 3; H, placenta 
with ovules [Harris 12081), x 20. Ardisia solanacea Roxb. ; I, dissected flower [Harris 
11637), calyx and ovary x i'5, corolla and stamens x 2. 



Ardisia byrsonimae is a slow-growing tree endemic to wooded limestone hill-tops 
in upper Clarendon, Jamaica. It differs from other Jamaican species in its compara- 
tively small leaves crowded at the ends of the shoots and its contracted inflorescences 
with very short pedicels. From A. tinifolia var. 3 {A. troy ana Urban), likewise 
inhabiting the western interior of Jamaica, it may also be distinguished by its 



i64 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

markedly imbricate and more rounded sepals. The epithet byrsonimae was given 
to this species by N. L. Britton under Icacorea but never published by him. It 
refers to a resemblance in habit between it and certain West Indian species of 
Byrsonima [Maipighiaceae], notably B. lucida (Mill.) Rich., with small obovate 
leaves usually clustered at the ends of the shoots. 

Subgenus 2. ICACOREA (Aublet) Pax 

Ardisia subgen. Icacorea (Aublet) Pax in Engler & Prantl, Nat. Pflanzenfam. 
4 (i): 94(1889). — Mez in Urban, Symb. Antill. 2 : 391(1901). 

Icacorea Aublet, Hist. PI. Guiane Franc. 2, Suppl. : i (1775). 

Ardisia sect. Icacorea (Aublet) Griseb., Fl. Brit. W, Ind. Is. : 395 (1861). 

Type (by monotjrpy): Ardisia guianensis (Aublet) Mez {Icacorea guianensis Aublet). 
Inflorescences terminal or both axillary and terminal; branches compoundly 
racemose, bearing several umbellate or corymbose clusters. Stamens with short 
filaments. Species all American. 

7. Ardisia compressa Kunth, Nov. Gen. & Sp. PI. 3 : 245 (1819). — Mez in Urban, 
Symb. Antill. 2 : 393 (1901); in Engler, Pflanzenr. IV. 236 : 89 (1902). — C. L. 
Lundell in Fieldiana, Bot. 24 (8) : 139 (1966). 

Ardisia decipiens sensu Griseb., Fl. Brit. W. Ind. Is. : 395 (1861) ; non A. DC. 
Icacorea compressa (Kunth) Standley in Contrib. U.S. Nat. Herb. 23 : 11 10 (1924). 

Hanover: Proctor 10013 (BM; IJ), Harris 10253 (BM; K; NY), Britton & Hollick 
2i2g (NY), Britton & Hollick 2164 (NY). 

Westmoreland : Proctor & Mullings 22060 (BM ; I J) . 

St. James: Proctor 16166 (BM; IJ). 

Ardisia compressa ranges from Mexico to northern South America, including 
Trinidad, but is not recorded from any other West Indian island. Humboldt and 
Bonpland collected the type near Caripe, Venezuela. Most continental specimens 
have the apex of the leaf more gradually tapered than have Jamaican specimens, 
which approximate most to some from British Honduras. The Jamaican specimens 
are all from the west of the island. The British Museum (Natural History) possesses 
an unlocalized specimen from Jamaica collected by William Wright who was in 
Jamaica in 1765-76 and 1782-85, living most of the time at Hampden, in Trelawny, 
and Mez suggests that this was of cultivated origin, which, however, seems unlikely. 
It is easily distinguished from other Jamaican species by its more intricately branched 
inflorescence, the lateral peduncles of which bear three or four secondary peduncles 
each with two to six small, whitish, brown-flecked flowers; the filaments of the 
stamens are about 0-5 mm. long, and much shorter than the anthers. 

Subgenus 3. TINUS Mez 

Ardisia subgen. Tinus Mez in Engler, Pflanzenr. IV. 236 : 124 (1902). 
Lectotype: Ardisia elliptica Thunb. {A. humilis sensu Mez, pro parte; non Vahl). 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 165 

Inflorescences axillary. Stamens with very short filaments. Species Asiatic and 
Australian, one often naturalized elsewhere. 

8. Ardisia solanacea Roxb., PI. Coast Coromand. i : 27, t. 27 (1795).— Sims in 
Curt. Bot. Mag. 40 : t. 1677 (1814).— Mez in Urban, Symb. Antill. 2 : 402 (1901); 
in Engler, Pflanzenr. IV. 236 : 132 (1902). — E. H. Walker in Philipp. Joum. 
Sci. 7Z ■ 58 (1940)- (Text-fig. 23 I.) 

Ardisia humilis sensu A.DC. in DC, Prodr. 8 : 129 (1844); non Vahl. — ^Wight, Ic. PI. Ind. 

Br. 4 : t. 1212 (1848).— Griseb., Fl. Brit. W. Ind. Is. : 396 (1861).— C. B. Clarke in Hook. 

f., Fl. Brit. Ind. 3 : 529 (1882). — De Wild., Ic. Select. Hort. Thenen. i : 79, t, 19 (1900). 
Icacorea solanacea (Roxb.) Britton, Fl. Barm. : 284 (19 18). 

St. James : Proctor 16461 (BM ; I J) . 

St. Mary: Harris 1163^ (NY; UCWI), Proctor 20613 (BM; IJ), Adams i2oy6 
(UCWI). 

St. Andrew: Grabham (NY), Adams ioog4 (BM). 

Portland: Powell yiy (BM; IJ; NY), Proctor 16546 (BM; IJ; NY), Proctor 19723 
(BM; I J; NY), Marble 852 (NY), Yuncker 18826 (BM), Adams 11482 (BM; UCWI). 

St. Thomas: Nichols igi (NY), Powell 854 (IJ), Harris & Britton 10585 (NY; 
UCWI), Proctor 92 ly (BM; IJ), Webster & Wilson 5218 (BM; IJ), Howard, Proctor & 
Steam 14809 (BM; IJ), Proctor 16450 (BM; IJ), Yuncker 17523 (BM), Yuncker 
18826 (BM; UCWI), Harris 6040 (NY), Britton 3555 (NY), Adams 10471 (UCWI). 

A native of India, Malaya and China, long cultivated in tropical gardens, this has 
become naturalized here and there in Jamaica, usually along streams or on shaded 
road banks, and is sometimes called " Blackberry " or " Craingcraing ". It is 
easily recognized by its axillary few-flowered inflorescences with comparatively 
large pale rose flowers clustered almost umbellately at the end of the long peduncle. 

2. WALLENIA Swartz 

Wallenia Swartz, Nov. Gen. & Sp. PI. : 2, 31 (1788); Fl. Ind. Occ. i : 247 (1797); 
nomen conservandum No. 6304. 

Petesioides Jacq. ex Kuntze, Revis. Gen. PI. 2 : 402 (1891), nom. superfl. 

Type (by monotypy) : W. laurifolia Swartz {Petesioides laurifolium Jacq.) 

Wallenia belongs to the tribe Myrsineae characterized by having the ovules few 
and arranged in a single series around the placenta. Both it and Myrsine can be 
readily distinguished by their small flowers, with the corolla 1-5-6 mm. long, from 
Ardisia of the tribe Ardisieae, characterized by having numerous ovules arranged in 
several series or apparently scattered over the placenta. The loose many-flowered 
inflorescence of Wallenia readily distinguishes it from Myrsine [Rapanea] which has 
flowers crowded in axillary clusters. The Jamaican species all belong to the sub- 
genus Wallenia (subgenus Euwallenia Mez) which is confined to Jamaica, Cuba and 
Hispaniola and has an erect terminal paniculate inflorescence; the other subgenus, 
Homowallenia Mez, with pendulous axillary racemose inflorescences, occurs in Cuba, 
Hispaniola, Puerto Rico, and the Lesser Antilles but not in Jamaica. 



i66 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

The dioecious nature of Wallenia possibly militates against the formation of well- 
defined taxa within the genus ; it certainly makes difficult the correlation of characters 
in herbarium material. The corolla of male flowers is much longer than the calyx; 
the anthers are borne on long exserted filaments. The corolla of female flowers is 
shorter or little longer than the calyx; the filaments are very short and the anthers 
aborted. Consequently female flowers are much less conspicuous than male flowers 
and are liable to be passed over by collectors as not being fully developed. Thus 
there is rarely available, from the same locality, male and female flowering material 
and mature fruiting material obviously belonging to the same species, which would 
provide the means of correlating other material. Instead some members of the 
genus are only known in one of these three possible states. The variability of the 
genus, presumably maintained by " out-breeding ", is such that two gatherings 
rarely match completely. Mez and Urban accordingly tended to describe each new 
gathering as representing a new species distinguished from its congeners by a com- 
bination of subtle rather than well-marked characters, the latter being few within 
the subgenus Wallenia {Euwallenia Mez) . The lack of comparable material makes it 
difficult to assess the worth of such characters. In preparing the present account it 
has not been possible to devise a key making full use of floral characters and unfortu- 
nately the vegetative characters used tend either to overlap or to be difficult to 
express without appearing too definite and exclusive or too vague and hence useless. 
Thus while the key should lead to the identification of typical material of the species, 
it would be too optimistic to expect conclusive results with all specimens. Wallenia 
laurifolia occurs in Cuba, Hispaniola and Jamaica, W. subverticillata in Cuba and 
Jamaica; the other species accepted here are apparently endemic to Jamaica. 

Key to Jamaican Species 

Leaves with base broadly rounded and slightly cordate, paired or in whorls of three 
together, fairly large (3-17 cm. broad), the petiole short (less than 6 mm. long) 
or lacking : 
Leaves rounded at the apex, 7-17 cm. broad: 

Leaves sessile . . . . . . . . . 1. W. clusioides 

Leaves with 3-6 mm. long petiole . . . . . 2. W. elliptica 

Leaves narrowed towards the acute or obtuse apex, 3-6 cm. broad 

3. W. subverticillata 
Leaves with base cuneate or attenuate, rarely rounded (the petiole then i cm. or 
more long), usually alternate, only occasionally paired or whorled, the petiole 
short or long : 
Leaves reticulate- veined beneath, with slightly raised cross-veins between the main 
lateral veins : 
Petiole short, c. 1-8 mm. long; blades under 11 cm. long: 

Leaves mostly broadest at the middle and gradually narrowed from there to 
the apex: 
Petiole very short, 1-4 mm. long; blade inconspicuously punctate beneath 

4. W. fawcettii 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 167 

Petiole longer (5-8 mm. or more long) ; blade conspicuously punctate 

beneath 10. 1^. venosa 

Leaves mostly broadest above the middle with the apex rounded or obtuse: 
Leaf-blades 2-8 cm. long, less than 3-5 cm. broad; crest of Blue Mountains 

5. W. crassifolia 
Leaf -blades larger (5-1 1 cm. long) : 

Leaves coriaceous, firm, greyish or glaucescent; female calyx c. 1-4 mm. 
long . . . . . . . . 6. W. corymbosa 

Leaves chartaceous, thinner: 

Leaves drying light green; female calyx c. 1-5 mm. long 

II. W^. purdieana 
Leaves drying brownish; female calyx i mm. long 

7. W. xylosteoides 

Petiole longer (c. 8-30 mm. long) ; blades mostly more than 11 cm. long: 

Leaf -blades nearly three times as long as broad, mostly narrowly elliptic : 

Base of blade narrowly cuneate or attenuate . . d>.W. erythrocarpa 

Base of blade rounded or abruptly contracted . . 12. W. grisebachii 

Leaf -blades about twice as long as broad, mostly elliptic or narrowly obovate : 

Lateral veins spreading outwards from the midrib at an angle of about 40° 

9. W. punctulata 
Lateral veins spreading outwards from the midrib at an angle of about 
50-70°: 
Male flower with calyx 1-5-2-5 mm. long, corolla 2-5-4 i^in- long; persis- 
tent fruiting calyx not more than 1-5 mm. long: 
Leaves usually densely blackish-punctate and somewhat mottled 
beneath, strongly veined, the base cuneate . 10. T^. venosa 

Leaves inconspicuously punctate and not mottled beneath, lightly 
veined, the base cuneate or rounded or abruptly contracted: 
Leaves drying light green ; inflorescence glabrous, loose ; male flower 
with corolla yellowish, not punctate; style about 0-8 mm. long 

II. T^. purdieana 

Leaves drying brownish; inflorescence minutely pubescent, compact; 

male flower with corolla reddish, distinctly punctate; style 

about 1-5 mm. long . . . . 12. T7. grisebachii 

Male flower with calyx c. 3 mm. long, corolla 5-6 mm. long; persistent 

fruiting calyx about 2-5 mm. long . . ^3- W. calyptrata 

Leaves pinnately veined beneath, i.e. without raised cross-veins between the main 

lateral veins : 

Blades of some or all leaves abruptly contracted at base into the petiole; 

inflorescence usually minutely pubescent . . . J2. W. grisebachii 

Blades of all leaves gradually narrowed into the petiole ; inflorescence glabrous : 

Blades of leaves mostly less than 12 cm. long, 6 cm. broad; petiole not more 

than I cm. long ...... 14. T^. laurifolia 

Blades of leaves up to 25 cm. long, 11 cm. broad, mostly more than 12 cm. 
long, 6 cm. broad; petiole 1-2 cm. long: 



i68 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

Fruits vertically ribbed . . . . . 15- W. sylvesiris 

Fruits not ribbed . . . . . . . 16. W. discolor 

1. Wallenia clusioides (Griseb.) Mez in Urban, Symb. Antill. 2 : 411 (1901); in 
Engler, Pflanzenr. IV. 236 : 245 (1902). 

Ardisia clusioides Griseb., Fl. Brit. W. Ind. Is. : 396 (1861) excl. specim. Alexander. 

Hanover: Harris 10333 (BM; NY), Proctor 10028 (IJ), Proctor 26590 (IJ). 

Westmoreland: Purdie (K, lectotype), Harris 10210 (BM; NY). 

St. James: Harris giyj (BM; NY). 

St. Elizabeth: Harris gy6i (NY), Britton mo (NY), Proctor 26530 (IJ). 

W. clusioides is a species of western Jamaica notable for its very large sessile and 
subcordate leaves 15-31 cm. long, 7-17 cm. broad. It forms a small tree or tall 
shrub 6-20 ft. (2-6 m.) high. The female flowers are unknown. The greenish- 
yellow male flowers have the calyx 1-5-2 mm. long, the corolla about 2-5 mm. long. 

2. Wallenia elliptica Urban, Symb. AntiU, 6 : 30 (1909). (PI. 9a.) 

Ardisia clusioides Griseb., Fl. Br. W. Ind. Is. : 396 (1861) pro parte, quoad specim. 
Alexander. 

St. Ann: Alexander 569 (K; NY; type-collection), Britton & Hollick 2yy6 (NY). 

An obscure species described by Urban on a specimen collected by Alexander at 
Grierfield near Moneague, St. Ann, W. elliptica is provisionally kept apart from 
W. clusioides on account of the shortly petioled leaves. The male flowers are 
unknown. The female flowers have the calyx and corolla about 1-5 mm. long. 

3. Wallenia subverticillata (Britton) Ekman ex Urban, Symb. Antill. 9 : 412 
(1925). — ^Alain in Leon & Alain, Fl. Cuba 4 : in (1957). (Text-fig. 24K.) 

Petesioides subverticillafum Britton in Bull. Torrey Bot. Club 37 : 355 (1910). 

Portland: Proctor 22102 (IJ). 

St. Thomas: Britton 393^ (NY), Harris & Britton 10694 (NY, holotype; UCWI), 
Harris & Britton loyoi (K; NY), Britton 4049 (NY), Proctor 28681 (BM). 

Wallenia subverticillata, a species of mountain woods at 1,000-2,000 ft. (300- 
600 m.), is distinguished by having its leaves grouped in pairs or threes; they are 
almost sessile and are much narrower and more attenuate towards the apex than 
those of W. clusioides but are nevertheless extremely variable in shape and size, 
e.g. 5-20 cm. long, 3-5 cm. broad on the same plant. The tjrpe was collected by 
Harris and Britton on a slope below Big Level, eastern St. Thomas. Rather 
surprisingly, specimens of what appear to be the same species have been collected 
in Las Villas, Cuba, by Ekman. The calyx of male flowers is 2-2-5 nii^i- long, of 
female flowers about i mm. long; the corolla of male flowers is about 2-5 mm. long, 
of female flowers about i mm. long, slightly longer than the calyx. 

4. Wallenia fawcettii Mez in Urban, Symb. Antill. 2 : 408 (1901); in Engler, 
Pflanzenr. IV. 236 : 244 (1902). (PI. 9b.) 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 169 

Portland: Maxon & Killip ^14 (NY). 

Portland or St. Thomas: Harris 5422 (NY, type collection). 

St. Thomas : Proctor 9684 (I J) . 

From other relatively small-leaved and short-petioled species, Wallenia fawcettii 
may be recognized by its more attenuate leaves, which are gradually narrowed above 
the middle to a usually acute apex. It is known only from the crest of the Blue 
Mountains, at about 5,000-5,500 ft. (1,500-1,700 m.) ; the type was collected by 
Harris near Portland Gap but whether on the Portland or St. Thomas side is un- 
certain. The calyx of male flowers is about 1-5 mm. long, of female flowers about 
I mm. long; the corolla of male flowers is about 2-5 mm. long, of female flowers 
unknown. 

5. Wallenia crassifolia Mez in Urban, Symb. Antill. 2 : 409 (1901); in Engler, 
Pflanzenr. IV. 236 : 244 (1902). (PI. loa.) 

Wallenia venosa Griseb., Fl. Brit. W. Ind. Is. : 394 (1861) pro parte, excl. specim. Wilson. 

St. Andrew: Proctor gSgo (IJ; NY), Proctor g$24 (BM; IJ), Maxon & Killip 1026 
(NY), Philipson 932 (BM), E. G. Britton 3866 (NY), Purdie (K, holotype). 

St. Thomas: Webster & Wilson 5445 (IJ), Proctor g2oy (IJ), Steam loi (BM), 
Proctor II 458 (IJ). 

Wallenia crassifolia inhabits mountain rain forest along the crest of the Blue 
Mountains between 4,750 and 6,500 ft. (1,400-2,000 m.). That the smallness of its 
elliptic or narrowly elliptic leaves is not simply a reduction occasioned by its high 
mountain environment is shown by the occurrence of the large-leaved and 
undoubtedly distinct W. calyptrata in the same habitat. The type-locality of W. 
crassifolia is Morces Gap, where it was first collected by William Purdie in 1843. 
The calyx of male flowers is about 2 mm. long, of female flowers about i mm. long, 
the corolla of male flowers about 4 mm. long, of female flowers unknown. 

6. Wallenia corymbosa Urban, Symb. Antifl. 5 : 457(1908). (Text-fig. 24A-J.) 
Trelawny: Harris 8y2o (BM; NY; UCWI; type-collection), Harris gog^ (NY; 

UCWI). 

Clarendon: Harris 10864 (NY), Harris iogg6 (NY), Harris I2y8i (BM; NY; 
UCWI), Harris 12803 (NY; UCWI). 

St. Ann: Howard & Proctor 13037 (IJ), Howard 6- Proctor 15170 (BM; IJ), Proctor 
26j3g (BM). 

Typical Wallenia corymbosa inhabits the central hills of Jamaica at about 2,000- 
2,500 ft. (600-760 m.), probably extending well into the Cockpit Country, and is 
notable for its thick greyish or glaucescent narrowly obovate leaves. 

The type-locality is on Crown Lands near Troy, where William Harris collected it 
at 2,500 ft. (760 m.) in 1904. Subsequent collections in the same area have provided 
an unusually adequate representation of it in male and female flower and in fruit. 
The flowers are described as " pale yellow with red spots " or " cream with red 



170 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

stripes ". The calyx of male flowers is 1*5-2 mm. long, of female flowers about 
1-4 mm. long, the corolla of male flowers 3-4 mm. long, of female flowers about 
2-5 mm. long. The fruits are about 3-5 mm. broad. 



Fig. 24. Wallenia corymbosa Urban: A, female flowering specimen {Harris 12803), x \\ 
B, female flower [Harris 12803), x 6; C, dissected calyx and gynoecium of female flower 
[Harris 12803), x 8; D, dissected corolla of female flower with aborted stamens [Harris 
12803), X 8; E, placenta of female flower [Harris 12803), x 20; F, male flower [Harris 
i28ig), X 8; G, dissected corolla of male flower with aborted gynoecium [Harris 12819), 
X 8; H, anther of male flower [Harris i28ig), x 20; I, fruit [Harris 12781), X 5; J, leaf 
[Harris 12781), x J. Wallenia subverticillata (Britton) Ekman: K, leaves [Harris &' 
Britton 4049), X J. Wallenia grisebachii Mez: L, leaf -base [Proctor 8247), x |. 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 171 

7. Wallenia xylosteoides (Griseb.) Mez in Urban, Symb. Antill. 2 : 409 (1901); 
in Engler, Pflanzenr. IV. 236 : 244 (1902). (PL lob.) 

Ardisia xylosteoides Griseb., Fl. Br. W. Ind. Is. : 395 (1861) tantum quoad pi. Jamaic. 

St. Ann: Alexander 568 (K, lectotype). 

Grisebach reported his Ardisia xylosteoides from " Haiti! ; Mexico! ; New Granada! " 
as well as Jamaica; these records cannot apply all to one species. Mez in 1901 
restricted the epithet xylosteoides to Jamaican material collected by Alexander on 
St. Ann's Road, Moneague. The same species is represented by Jamaican Plants 
1118 (UCWI) and 1141 (K) without locality. The calyx of female flowers is about 
I mm. long, the corolla scarcely longer. Male flowers and fruit are unknown. The 
epithet indicates a resemblance in leaf-form to Lonicera xylosteum L., the Xylosteum 
of pre-Linnaean authors. 

8. Wallenia erythrocarpa Urban, Symb. Antill. 6 : 29 (1909). (PI. iia.) 

Hanover: Harris 10343 (BM; NY; UCWI; type-collection). 

Westmoreland: Harris yoSo (UCWI). 

This species is still only known from material collected by Harris, the type from 
Fray Woods at 1,650-1,800 ft. (500-550 m.) in 1908; he gathered other material at 
Grandvale at 500 ft. (150 m.) in 1898. It is notable for its large but relatively narrow 
long-petioled leaves with prominent reticulate veining, their narrowly elliptic-oblong 
shape and their size (11-21 cm. long, 4-5-7-5 cm. broad) distinguishing it from W. 
venosa. The calyx in fruit is about i mm. long. Male and female flowers are un- 
known. 

9. Wallenia punctulata Urban, Symb. Antill. 7 : 322 (1912). (PI. lib.) 

Manchester: Britton 3281 (NY; type collection). 

Wallenia punctulata is another little-known species, known only from the type- 
collection gathered by N. L. Britton in 1908 between Brown's Town and Porus, but 
distinguished from other species with prominently reticulate-veined leaves by its 
more oblique lateral veins which pass outwards from the midrib at an angle of about 
40°. The calyx in fruit is about 2 mm. long, the corolla of female flowers according 
to Urban about 1-7 mm. long. 

10. Wallenia venosa Griseb., Fl. Brit. W. Ind. Is. : 394 (1861) quoad specim. 
Wilson. — Mez in Urban, Symb. Antill. 2 : 410 (1901); in Engler, Pflanzenr. IV. 
236 : 245 (1902). — Urban, Symb. Antill. 7 : 322 (1912). (PI. 13a.) 

Hanover: Britton 2425 (NY), Britton 22g2 (NY), Britton 2326 (NY). 

Westmoreland : Harris 10248 (BM ; NY) . 

Trelawny : Harris 8768 (BM ; NY ; UCWI) . 

St. Elizabeth: Harris 9925 (BM; K; NY; UCWI). 

Manchester: Britton 3712 (NY). 

Clarendon: Proctor 18400 (BM; IJ), Proctor 26514 (BM; IJ), Proctor 27637 (BM; 

IJ). 
St. Ann : Proctor 26514 (BM) . 



172 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

St. Andrew: Harris 5633 (BM; NY; UCWI), Harris 6126 (BM), Harris 16032 
(BM; NY; UCWI), Adams 11845 (UCWI). 

St. Thomas: Harris & Britton 10745 (NY), Britton 3g63 (NY), Britton 3g88 (NY), 
Britton 4145 (NY) . 

The lectotype of this widespread species is Wilson 388 (GOET, K), collected in 
1858 without specified locality, which was designated by Mez in 1901. Wallenia 
venosa has prominently reticulate-veined leaves with black dots beneath sometimes 
so numerous as to give it a somewhat mottled appearance. The calyx of male 
flowers is 1-5-2 mm. long, of female flowers o-8 mm. long, the corolla of male flowers 
about 3 mm. long, of female flowers about i mm. long. The Alexander specimen 
(GOET) from the Blue Mountains cited by Grisebach under W. venosa belongs to 
W. crassi folia. 

11. Wallenia purdieana Mez in Urban, Symb. Antill. 2 : 408 (1901); in Engler, 
Pflanzenr. IV. 236 : 243 (1902), (PI. 12a.) 

Hanover: Harris 10305 (BM; UCWI), Proctor 11301 (IJ). 

St. James: Proctor 22gg2 (BM; IJ). 

Trelawny: Harris gog6 (UCWI), Britton 510 (NY), Proctor 13663 (BM; IJ), 
Howard, Proctor & Steam 14663 (BM; IJ). 

Manchester: Purdie (K, type), Proctor 16118 (BM; IJ), Howard, Proctor & 
Wagenknecht 20520 (NY), Harris & Britton 10601 (BM; NY; UCWI), Robertson 
5416 (UCWI). 

Wallenia purdieana is a species of western and central Jamaica between 1,000 ft. 
(300 m.) and 3,000 ft. (900 m.). The type was collected by William Purdie in 
Manchester, but without specified locality. The leaves on drying retain their green 
colour unlike those of other species which become brownish or grey. The calyx of 
male flowers is about 2 mm. long, of female flowers about 1-5 mm. long, the corolla 
of male flowers about 3 mm. long, of female flowers about 1-5-2 mm. long. 

12. Wallenia grisebachii Mez in Urban, Symb. Antill. 2 : 411 (1901) quoad specim. 
WuUschlagel 1338; in Engler, Pflanzenr. IV. 236 : 245 (1902). — ^Urban, Symb. 
Antill. 6 : 31 (1909). (Text-fig. 24L; PI. 12b.) 

Wallenia laurifolia sensu Griseb., Fl. Brit. W. Ind. Is. : 394 (1861) pro parte; non Swartz. 

St. James: Proctor 22154 (BM; IJ). 

St. Elizabeth: Howard & Proctor 13745 (IJ; NY), Howard & Proctor 13755 (IJ). 

Trelawny: Perkins 1385 (K). 

Manchester: WuUschlagel 1338 (Munich, lectotype), S. Brown 300 (NY), Robertson 
5470 (UCWI), Adams 10116 (UCWI), Adams 11766 (UCWI). 

Clarendon: Proctor 8247 (IJ). 

St. Ann: Britton & Hollick 2776 (NY; UCWI). 

The species W. grisebachii as accepted here consists of plants with fairly large 
coriaceous leaves having the rounded base contracted abruptly into the petiole and 
becoming brownish on drying. Mez separated W. grisebachii from among material 
included by Grisebach in W. laurifolia. Urban further divided this, restricting the 
name W. grisebachii to the taxon represented by WuUschlagel 1338 collected at 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 173 

Fairfield, Springfield, in Manchester parish, between 1847 ^^^ 1849. The inflores- 
cence is compact; the flowers are rusty brown. The calyx of male flowers is 2 mm. 
long, of female flowers 1-5-2 mm. long, the corolla of male flowers about 4 mm. long, 
of female flowers about 1-5 mm. long. 

13. Wallenia calyptrata Urban, Symb. Antill. 5 : 458 (1908); in Fedde, Repert. 
Sp. Nov. 13 : 469 (1915). 

St. Andrew: Shreve s.n. (NY), /. A. Harris 6- Lawrence 15490 (NY), Adams ^440 
(UCWI). 

Portland: Proctor 26yiy (BM; IJ). 

St. Thomas: Proctor 9638 (IJ), Proctor 9683 (IJ; NY), Steam 104 (BM), Proctor 
1 1457 (IJ). 

Although the type-gathering of W. calyptrata made by Rehder in 1903 at Morces 
Gap has not been seen, the agreement of Urban's description with the other material 
collected along the crest of the Blue Mountains leaves no doubt as to its identity. 
Its range here seems to be between 5,000 ft. (1,500 m.) and 7,000 ft. (2,100 m.) in 
mountain mist-forest. It has prominently reticulate-veined leaves 8-14 cm. long, 
3-5-6 cm. broad, and larger male flowers than other species. The calyx of male 
flowers is 3-3-5 mm. long, of female flowers 2-5 mm. long, the corolla of male flowers 
5-6 mm. long, of female flowers 3 mm. long. 

14. Wallenia laurifolia Swartz, Nov. Gen. & Sp. PI. : 31 (1788); Fl. Ind. Occ. 
1:248 (1797). — ^Mez in Urban, Symb. Antill. 2 : 407 (1901); in Engler, 
Pflanzenr. IV. 236 : 243 (1902). — Urban, Symb. Antill. 8 : 521 (1921). — Alain 
in Leon & Alain, Fl. Cuba 4 : no, fig. 42 (1957). 

Petesioides laurifolium Jacq., Select. Stirp. Amer. Hist. : 17 (1763), nom. invalid. 
Wallenia clusiifolia Griseb., Fl. Brit. W. Ind. Is. : 394 (1861). 

" Bryonia nigra fruticosa, foliis laurinis, floribus racemosis speciosis " Sloane, Cat. PL Jam. : 
106 (1696); Voy. Jam. Nat. Hist, i : 234, t. 145 f. 2 (1707). 

Hanover : Proctor 10035 (BM ; I J) . 

Manchester: 5. Brown 138 (NY), Proctor 24175 (BM; IJ), Crawford 775 (NY), 
Farr & Sanderson 22334 (IJ), Harris & Britton 10613 (NY; UCWI), Howard & 
Proctor 14964 (IJ), Howard, Proctor & Wagenknecht 20520 (BM; IJ), Harris 12862 
(BM; NY), Howard, Proctor & Steam 14708 (BM; IJ). 

Clarendon: Harris 11188 (NY; UCWI). 

St. Ann: Steam 603 (BM), Proctor 11907 (IJ), Proctor 8632 (IJ), Proctor 16142 (IJ). 

rST. Catherine: Britton 2616 (NY). 
St. Mary: Sloane, Herb. IV. 494 (BM), Proctor 7554 (NY; IJ). 
St. Andrew : Harris 6880 (BM ; NY ; UCWI) . 
* Portland: Steam 538 (BM), Proctor 11841 (IJ), Wight 195 (NY), Maxon & 
Killip 830 (NY), Harris & Britton 10767 (K; NY), Proctor 11347 (BM; IJ), Proctor 
11352 (IJ), Proctor 16259 (IJ), Howard, Proctor & Steam 14758 (BM; IJ), Steam 226 
(BM; UCWI), Britton 4121 (NY). 
St. Thomas: Britton 4109 (NY), Powell 1647 (BM; IJ). 



174 A SYNOPSIS OF JAMAICAN MYRSINACEAE 

This widespread species, first described from Hispaniola and also known from 
Cuba, is mainly distinguished by its relatively inconspicuous veining with the areas 
between the main lateral veins flat, not reticulate with prominent veinlets. The 
material from the upper part of the John Crow Mountains exemplified by Proctor g822 
gathered at 1,500-2,500 ft. (450-750 m.), with very small leaves (to 6 cm. long) and 
reduced inflorescence, differs much in appearance from that of the coast, exemplified 
by Steam 226, but these differences are assumed to result from the diversity of 
habitat, as small leaves approaching these are to be found on some specimens from 
elsewhere referred to this species, e.g. Howard, Proctor & Steam 14^08 from 
Manchester, Proctor 8632 from St. Ann, and Proctor 10035 from Hanover, and inter- 
mediates occur moreover on the John Crow Mountains. The calyx of male flowers 
is I-5-2-5 mm. long, of female flowers about 1-5 mm. long, the corolla of male flowers 
2-5-3 mm. long, of female flowers about 1-5 mm. long. 

15. Wallenia sylvestris Urban in Fedde, Repert. Sp. Nov. 13 : 468 (1915). 
(PI. 13b.) 

Portland: Proctor i04yy (IJ; NY), Proctor 16258 (IJ), Proctor 5266 (BM; IJ), 
Proctor 5244 (IJ). 

Portland or St. Thomas: Harris & Britton ioy2o (NY, type-collection), Harris & 
Britton io6go (K; NY; UCWI), Harris & Britton 10776 (NY; UCWI). 

This species is apparently confined to mountain woods on the John Crow Mountains 
at 1,500-2,500 ft. (450-750 m.). The type-locality is the southern end of the John 
Crow Mountains near the boundary of the parishes of Portland and St. Thomas. It 
is notable for its large long-petioled leaves with elliptic or narrowly elliptic (rarely 
obovate or narrowly obovate) blades up to 22 cm. long, 9 cm. broad. The calyx and 
corolla of female flowers are about 2 mm. long. The male flowers are unknown. 
The subglobose fruits are slightly punctate and vertically ribbed. 

16. Wallenia discolor Urban, Symb. Antill. 6 : 29 (1909). 

St. Ann: Alexander (fide Urban). 

Described by Urban from fruiting material collected by Alexander, probably in 
St. Ann but without recorded locality, this appears to be distinguished from W. 
laurifolia by its larger leaves (18-25 cm. long), shorter pedicels (0-5-3 mm. long) and 
coarsely punctate but not ribbed fruits. Male and female flowers are unknown. 

3. MYRSINE L. 

Myrsine L., Sp. PL i: 196(1753), Gen. PI. ed. 5: 90(1754). 
T5rpe (by monot3rpy) : M. africana L. 

The genus Myrsine as accepted here includes the groups kept apart by Mez as 
Myrsine, Rapanea and Suttonia. 

Key to Jamaican Species 
Young shoots and petioles minutely pubescent ; leaf-blades mostly narrowly elliptic 
and mostly less than 2-5 cm. broad; corolla lobes joined for the lower 0-5 mm. 
or more . . . . . . . . . . 1. M. coriacea 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 175 

Young shoots and petioles glabrous ; leaf -blades mostly obovate or narrowly obovate 
and mostly more than 2-5 cm. broad; corolla lobes free to the base 

2. M. acrantha 

I. Myrsine coriacea (Swartz) R. Br. ex Roem. & Schult. in L., Syst. Veg., ed. 
nov. 4 : 511 (1819).— A. DC. in DC, Prodr. 8 : 99 (1844). (Text-fig. 25E-H.) 

Samara coriacea Swartz, Nov. Gen. & Sp. PI. : 32 (1788); Fl. Ind. Occ. i : 261 (1797). 

Caballeria ferruginea Ruiz & Pavon, Syst. Veg. Fl. Peruv. & Chil. i : 280 (1798). 

Myrsine ferrugine a (Ruiz & Pavon) Spreng. in L., Syst. Veg., ed. 16, i : 664 (1825). 

Myrsine laeta sensu Griseb., Fl. Brit. W. Ind. Is. : 392 (1861); non A.DC. 

Rapanea coriacea (Swartz) Mez in Urban, Symb. Antill. 2 : 428 (1901) ; in Engler, Pflanzenr. 
IV. 236 : 380 (1902). 

Rapanea ferruginea (Ruiz & Pavon) Mez in Urban, Symb. Antill. 2 : 429 (1901); in Engler, 
Pflanzenr. IV. 236 : 381 (1902). — Urban, Symb. Antill. 8 : 522 (1921). — Britton & Wilson 
in Sci. Surv. Puerto Rico 6 : 61 (1925).- — Alain in Le6n & Alain, Fl. Cuba 4 : 112, fig. 43 
(1957).— J. F. Macbr. in Publ. Field Mus. Nat. Hist., Bot. 13 (5) : 168 (1959).— Little 
& Wadsworth, Common Trees of Puerto Rico : 432, fig. 203 (1964). 

St. James : Norman 50 (BM) . 

Trelawny: Webster, Proctor & Powell 5355 (BM; IJ), Britton 486 (NY), Adams 
12413 (UCWI). 

Manchester: Proctor 2ig2o (BM; IJ), Britton 3216 (NY), Adams 8466 (UCWI), 
Proctor 2 31 18 (BM; IJ). 

Clarendon: Proctor gy23 (IJ), Proctor 842y (IJ), Proctor 15905 (BM; IJ; UCWI). 

St. Ann: Proctor & Howard 14992 (IJ), Howard, Proctor & Steam 1460^ (BM; IJ), 
Proctor 7471 (BM; IJ), Skelding 3016 (UCWI), Adams 12698 (UCWI). 

St. Catherine: Howard & Proctor 13599 (U)- 

St. Andrew: Adams 12675 (UCWI), Britton 147 (NY). 

Portland: Philipson 933 (BM), Proctor loiio (IJ; NY), Howard, Proctor & 
Steam 14767 (BM; IJ). 

St. Thomas: Webster & Wilson 5454 (IJ), Steam iii (BM), Proctor 11463 (IJ), 
Maxon 9442 (US), Proctor 1108 (IJ), Steam 505 (BM). 

A widely ranging species, recorded in the West Indies from Cuba, Hispaniola, 
Jamaica, Puerto Rico and the Lesser Antilles and on the American mainland from 
Mexico to Peru, Chile, Bolivia and Argentina, this is commonly known as Rapanea 
ferruginea, a name based on Peruvian material; Swartz's Samara coriacea, based on 
Jamaican material being, in my opinion, conspecific, his epithet is adopted here. 
It is sometimes called " colic wood " in Jamaica, according to Proctor. 



2, Myrsine acrantha Krug & Urban in Notizbl. K. Bot. Gart. Mus. Berlin 1 : 79 
(1895); in Urban, Symb. Antill. i : 380 (1899). (Text-fig. 25A-D.) 

Myrsine coriacea sensu Griseb., Fl. Brit. W. Ind. Is. : 392 (1861); non (Swartz) R. Br. 
Rapanea guianensis Mez in Urban, Symb. Antill. 2 : 431 (1901); in Engler, Pflanzenr. IV. 

236 : 392 (1902) pro parte, quoad pi. Jamaic. ; non Aubl. 
Rapanea acrantha (Krug & Urban) Mez in Urban, Symb. Antill. 2 : 433 (1901) ; in Engler, 

Pflanzenr. IV. 236 : 376 (1902). 



176 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 




Fig. 25. Myrsine acrantha Krug & Urban: A, fruiting specimen (Robertson 5372), x i; 
B, flower [Harris 8295), x 10; C, dissected flower with aborted gynoecium (Harris 82g5), 
X 10; D, fruit (Harris yo6i), x 5. Myrsine coriacea Swartz: E, leaf (Steam iii), X ^; 
F, leaf (Steam 505), x J; G, dissected corolla with stamens (Steam iii), X 10; H, 
dissected calyx with gynoecium (Steam iii), X 10. 



Westmoreland : Proctor & Mullings 22061 (BM ; I J) . 

St. Elizabeth: Britton 1320 (NY), Proctor 7754 (BM; IJ), Howard & Proctor 
13682 (IJ). 

Manchester: Howard, Proctor & Wagenknecht 20521 (BM; NY; UCWI), Robertson 
5372 (BM; K; UCWI), Harris 8295 (BM; NY; UCWI), Proctor 2igo8 (BM; IJ), 
Proctor ii6ig (BM; IJ), S. Brown 159 (NY), Adams 10114 (UCWI), Proctor 23130 

(U). 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 177 

St. Andrew: Harris 6oy2 (NY), Harris SjgS (K; NY; UCWI; type-collection of 
Myrsine acrantha), Harris ^jgS bis (BM; UCWI), A. von der Porten 2035 (U)' Harris 
.5528 (NY; UCWI), Loveless 1502 (IJ; UCWI). 

Portland: Proctor igysy (BM; NY; UCWI). 

St. Thomas: Proctor 9580 (BM; IJ). 

Myrsine acrantha has hitherto been regarded as a rare and obscure species, known 
only from the type and topotype collections gathered by Harris in 1894 and 1895 
at Strawberry Hill Cottage high in the Blue Mountains of St. Andrew. Mez distin- 
guished it by its free petals from the two other Jamaican species accepted by him 
under the names Rapanea ferruginea and R. guianensis. Thus according to his own 
generic key in the Pflanzenreich IV. 236 : 15 (1902) it should be referred to the 
otherwise Pacific Ocean group Suttonial In fact all the Jamaican specimens hitherto 
referred to Myrsine (or Rapanea) guianensis agree with M. acrantha in having obovate 
or narrowly obovate leaves, flowers borne on very short bract-covered lateral shoots, 
and free petals. Myrsine guianensis (Aubl.) Kuntze does not occur in Jamaica; it 
was described by Aublet in 1775 (Hist. PI. Guiane Fran9. i : 121 ; 3 : t. 46) as Rapanea 
guianensis from French Guiana and is accepted by Mez as ranging from Florida over 
the West Indies to the Guianas and over Mexico and Central America to Bolivia. 
Tjrpical M. guianensis as represented by specimens from French Guiana {Aublet! 
Sagot gij! Martin/) is notable for its large obovate leaves up to 10 cm. long, 7 cm. 
broad, with a glossy almost varnished upper surface and distinctly punctate lower 
surface; plants of similar character occur in Bolivia, Colombia, British Guiana, and 
Surinam. Very similar to this but distinguishable by the leaves being marked with 
short lines beneath is M. trinitatis A. DC. occurring in Trinidad, Tobago, Grenada, 
Martinique and Puerto Rico. Plants from Hispaniola, Cuba, Jamaica, the Bahamas 
and Florida stand apart from these in their smaller leaves, rarely more than 8 cm. 
long, 2-5 cm. broad, less glossy above, punctate but not lined beneath. Whether 
more than one taxon is represented in this mostly sterile material is uncertain. 
Some specimens from Cuba, e.g. Ekman 15603 (Oriente province), and from Haiti, 
e.g. E. C. Leonard 8456 (Departement du Nord) in their leaves and short spurs 
unfortunately lacking flowers, closely resemble the Jamaican M. acrantha. In the 
Bahamas and Florida the plants of this alliance as represented by A . Curtiss 44 from 
New Providence, Bahamas, and R. A. Howard 804^ and Tracy 7454 from Florida 
have their five corolla-lobes fused at the base for about 0-4-0 -5 mm., as figured by 
Small (Man. Southeast Fl. : 1029 (1933)). They represent a species distinct from 
M. acrantha for which the correct name appears to be Myrsine punctata.^ 

* Myrsine punctata (Lam.) Steam, comb. nov. 

Sideroxylum punctatum l^duxa.., Tahl. 'Encyc. &.M.Qth..,'Bot. 2: 42 (1794) [from Carolina, Michaux]. 

Bumelia punctata (Lam.) Roem. & Schult. in L., Syst. Veg., ed. nov. 4 : 498 (1819). 

Myrsine floridana A. DC. in Trans. Linn. Soc. London 17 : 107 (1834); in DC, Prodr. 8 : 98 (1844) 

[from Florida]. 
Rapanea guyanensis sensu Small, Man. Southeast Fl. : 1029, fig. (1933); non Aublet. 

SPECIAL LITERATURE 
Grisebach, a. H. K. 1859-64. Flora of the British West Indian Islands. London. 
HosAKA, E. Y. 1940. A revision of the Hawaiian Species of Myrsine (Suttonia, Rapanea), 
Myrsinaceae. Occ. Pap. Bernice P. Bishop Mus., Honoluhi. 16 : 25-79. 



178 



A SYNOPSIS OF JAMAICAN MYRSINACEAE 



LuNDELL, C. L. 1966a. Flora of Guatemala: Myrsinaceae. Fieldiana, Bot. 24 (8) : 135-200. 
1966&. The genus Parathesis of the Myrsinaceae. Contrib. Texas Res. Found. Bot. Studies, 

5. Renner, Texas. 
Mez, C. 1901. Myrsinaceae. Urban, Symb. Antill. 2 : 389-433. 

1902. Myrsinaceae. Engler, Pflanzenreich IV. 236 (Myrsinaceae). 

Stearn, W. T. 1965. Grisebach's Flora of the British West Indian Islands: a biographical 

and bibliographical introduction. Journ. Arnold Arb. 46 : 243-285. 
SwARTZ, O. 1797. Flora Indiae occidentalis i : 247-249, 468-469. 
Urban, I. 1898. Bibliographia Indiae occidentalis botanica. Urban, S^/mfe. ^w^i//. i : 3-195. 

1909. Incrementa siphonogamarum florae jamaicensis. Urban, Symb. Antill. 6 : 70-131. 

Walker, E. H. 1940. A revision of Eastern Asiatic Myrsinaceae. Philipp, Journ. Sci. 

73 : 1-258. 
1954- Concerning the Myrsinaceae (" Ardisiaceae ") of Japan. Bot. Mag. Tokyo 67 : 

105-111, 155-162, 203-213, 247-255. 




Dr. William T. Stearn, D.Sc, Sc.D. 
Dept. of Botany 

British Museum (Natural History) 
Cromwell Road, London, S.W.7 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE6 




ARDISlATlnifolia 



Ardisia tinifolia Swartz; original drawing by Swartz. 



BOT. 4, 4 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE 7 




ARDISIA coria.cea. 



Ardisia tinifolia Swartz; original drawing by Swartz representing A. coriacea Swartz. 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



V / 



^•T 







PLATE 8 


mm 


^ 


r *■ 

-. ;; .,' i-^ i 


•j 


i 


5 1 i^! 

1 -*" " " • 


^^^^ 


tj^ 




m 




.^ 


P- 


^"'^^ "^-"^^-^ 






I \ 




► 









^ 



a, 



^ 
g 




t>0 

10 






\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 

fi 



PLATE9 



i 







J 




.-^ 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE 10 



I 








O 14J 

O CO 

^ .5 
ft 

^•^ 
■^ en 



? > 

■^ o 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE II 




00 



o 



1 




.^ 



Oh 






is 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE 12 




00 




ih 



si 




s 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 4 



PLATE 13 








00 



o 



O 



\ 



BY ADLARD & SON LIMITED 
BARTHOLOMEW PRESS, DORKING 



THE JAMAICAN SPECIES OF 

COLUMNEA AND ALLOPLECTUS 

(GESNERIACEAE) 



WILLIAM T. STEARN 




BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

lOTANY Vol. 4 No. 5 

LONDON : 1969 



THE JAMAICAN SPECIES OF COLUMNEA AND 
ALLOPLECTUS (GESNERIACEAE) 



BY 



WILLIAM THOMAS STEARN / 



British Museum (Natural History) 



Pp. 179-236; 29 Text-figures, Plates 14-21 




BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No, 5 

LONDON 11969 



THE BULLETIN OF THE BRITISH MUSEUM 

(NATURAL HISTORY), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 5 of the Botany series. 



World List abbreviation 
Bull. Br. Mus. nat. Hist. (Bot.). 



Trustees of the British Museum (Natural History) 1969 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

Issued 10 June, 1969 P^ice £1 8s. 



\ 



THE JAMAICAN SPECIES OF COLUMNEA AND 
ALLOPLECTUS (GESNERIACEAE) 

By WILLIAM T. STEARN 

With Notes on CYTOLOGY and CHEMOTAXONOMY 

By BRIAN D. MORLEY 

SUMMARY 

Eleven species of Columnea L., together with hybrids between two species (C. ruHlans and 
C. urbanii), and two species of Alloplectus Mart, occur in Jamaica. All are endemic and have 
distinctive ranges within the island. Columnea proctorii and C. urbanii, here described as new 
species, have previously been misidentified. 

The present paper deals with their history and morphological characteristics, the typification 
of the names Columnea, Achimenes and Alloplectus, the definition of the genera Columnea and 
A lloplectus, and the effect of pollination by humming-birds on their evolution ; it also provides 
notes on cytology and chemotaxonomy as well as a key, descriptions and enumerations of speci- 
mens. Diagnostic features of pubescence and calyx are illustrated. A survey by methods of 
numerical taxonomy (taxonometrics) using both a wide range of characters and a reduced range 
gave a classification of the species in general agreement with one reached by traditional methods. 

Introduction 

The following paper on the Jamaican species of Columnea and Alloplectus is the out- 
come of the preparation of the text on the Gesneriaceae for Fawcett and Rendle's 
Flora of Jamaica, volume 6. A study of all the available herbarium material before 
my visit to Jamaica in 1955-56 (cf. Stearn, 1959) indicated the need for further 
collecting and study of the species, despite the excellence of Morton's careful and 
critical revisions of 1944, and I gathered specimens of them whenever I came across 
them. Thanks to the co-operation of Mr. George R. Proctor of the Institute of 
Jamaica I was able to see most of them in their type localities or nearby. Subsequent 
collecting by Proctor has further increased our knowledge of their distribution. In 
1965 Dr. Brian Morley visited Jamaica to study the cytology of the genus Columnea, 
using wild material. His field work revealed the existence of hybridization between 
C. rutilans and C. urbanii and he was able to introduce most of the species into 
cultivation at the Royal Botanic Gardens, Kew. Study of the abundant material 
now available made evident the need also to consider the definition of the genera 
Columnea and Alloplectus on a wider basis. It provided data which appeared suitable 
for testing taxonometric analysis as a means of classifying the species. These 
matters are dealt with below. 



The type of columnea 

The genus Columnea L. (sensu stricto) comprises more than 100 species, all native 
to tropical America. It attains its greatest morphological diversity and breaks up 

BOX. 4, 5. II 



i82 THE JAMAICAN SPECIES 0¥ COLUMNEA K^T> A LLOPLECT U S 

into its greatest number of species in northern South America and Central America : 
17 species, for example, are recorded from Costa Rica alone. In the West Indies as 
a whole it is represented by some 16 species, one of which, C. scandens L., is the 
type-species by monotypy. Charles Plumier collected this between 1689 and 1697 
" in locis insulae martinicanae ", i.e. in Martinique, Lesser Antilles, and founded a 
new genus upon it dedicated to the Italian botanist Fabius Columna (i.e. Fabio 
Colonna, 1567-1650). He published a brief description, together with an engraving 
showing floral details, in his Nova Plantarum Americanarum Genera : 28, t. 33 (1703) 
as follows: 

" Columnea est plantae genus flore monopetalo, personato, cujus labium superius 
non nihil fornicatum & excavatum, inferius vero tripartitum. Ex calyce autem 
surgit pistillum, posticae floris parti, ad instar clavi infixum, quod deinde abit in 
fructum globosum, mollem, seminibusque plenum exiguis, & oblongis. 

Columneae species sunt. 
Columnea scandens, phaeniceo flore, fructu albo, 
Columnea scandens, flore lutescente, fructu albo." 

Plumier had, however, drawn up a much more detailed description and had made 
a coloured drawing depicting the flowers as red, the original of which is at the 
Museum National d'Histoire Naturelle, Paris, but of which there are copies in the 
University Library, Groningen, the Herbarium Library of the Royal Botanic 
Gardens, Kew (Plum. MSS. 2 : 117), and the Department of Botany, British Museum 
(Natural History), London (Plum. MSS. 2 : 104). That now at Groningen forms part 
of the set of copies of Plumier's drawings originally made for Boerhaave, which 
passed after his death to Johannes Burman, Linnaeus's friend, at Amsterdam. In 
1756 Burman published an engraving made from it, excellently portraying the habit 
of the whole plant, in Plantarum Americanarum Fasciculus quartus: t. 89. 
Lamarck's account in the Encyclopedie methodique, Botanique 2 : 66 (1786), which is 
fuller than any previously published, was based on Plumier's manuscript material 
in Paris. Plumier's plant with yellowish flowers is assumed to be a rare colour 
form. 

Linnaeus himself possessed no specimens of Columnea and he based his description 
of the genus in his Genera Plantarum: 373 no. 921 (1737), 5th ed. : 286 no. 710 (1754) 
and likewise his protologue of C. scandens in the Species Plantarum 2 : 638 (1753) 
primarily on Plumier's published text and figures, although it is evident that between 
the preparation of the first edition (1737) and the second edition (1742) of the Genera 
Plantarum he saw herbarium material of a Columnea which enabled him to amend 
his 1737 statement " STAM: Filamenta duo (pingit, an quatuor) ", based on Plumier's 
figure which portrays only two stamens, to " STAM. Filamenta quatuor, quorum duo 
longiora, sub labio superiore recondita. Anther ae simplices " in 1742. He may well 
have seen material during his visit to Paris in 1738. In any event he stated " Habitat 
in Gallia aequinoctiali ", meaning the French possessions in tropical America. The 
traditional use of the name Columnea scandens L. for the red-flowered species 
occurring in Martinique is thus well-founded. This ranges in the Lesser Antilles 
from Montserrat, by way of Martinique, St. Vincent and Grenada, to Trinidad and 
Tobago, but does not occur in the Greater Antilles. 



the jamaican species ofcolf/mtvcj and a llop lect us 183 

Sloane's Jamaican material 

The first record of any species now included in the genus Columnea was made not 
by Plumier but by Hans Sloane and resulted from his stay in Jamaica from December 
1687 to March 1689. He found " in montosis Insulae Jamaicae sylvis " a species 
which he listed as " Rapunculus fruticosus foliis oblongis integris villosis ex adverso 
sitis flore purpureo villoso " in his Catalogus Plantarum quae in Insula Jamaica 
sponte proveniunt : 58 (1696). This entry would be completely cryptic but for the 
description and the rather crude engraving in his later Voyage to the Islands Madera . . . 
and Jamaica i : 157, t. 100 f. i (1707) and the specimen in Herb. Sloane 3 : fol. 21 
at the British Museum (Natural History) on which these were based. Sloane's artist 
Everard Kickius unfortunately failed to distinguish the calyx from the corolla, 
both being very hairy, but examination of Sloane's specimen shows his plant to be 
Columnea fawcettii (Urban) Morton, which was not distinguished as a species until 
1944. Sloane probably collected it on Mount Diablo. 

The type of Achimenes 

Patrick Browne, failing to connect Linnaeus's genus Columnea with any Jamaican 
plant, established a new genns Achimenes in his Civil and Natural History of Jamaica : 
270, t. 30 (1756). He included two species. The first of these, his " Achimenes 
major, herbacea, subhirsuta, oblique assurgens " to which he referred Sloane's 
" Rapunculus fruticosus . . ." is undoubtedly a Columnea, probably C. fawcettii, 
since he mentions it as " most commonly met with in the woods of New Liguanea 
and St. Ann's " and describes the " divisions of the cup ", i.e. calyx-segments, as 
" pinnated at the sides, somewhat like those of the garden rose ". Browne's generic 
description was based on this species, as is evident, for example, from his account of 
the irregular corolla: " limbus erectus, in quatuor lacinias inaequales, inaequaliter 
sectus; lacinia superior recta, latiuscula, profunde crenata, sive bifida; laterales 
oblongae & a superiori oblique decendentes; infima anguste patula, ad medietatem 
floris incisa, longissima." Hence Achimenes P. Browne (1756) as to type is congeneric 
with Columnea L. Browne's second species " Achimenes minor, erecta, simplex ..." 
is the plant commonly known as Achimenes coccinea (Scop.) Pers. or A. pulchella 
(L'Herit.) Hitchc, but whose correct name is A. erecta (Lam.) H. P. Fuchs. Browne 
regarded it as agreeing " in the most essential parts [of the flower] with the fore- 
going ", although it has an almost regular corolla. When Persoon dealt with these 
two Brownean species in his Synopsis Plantarum 2 : 164 (1806) he referred Browne's 
first species to Columnea, as Burman had already done in 1756 and Swartz in 1788, 
but he retained the na.me Achimenes for the second species, at the same time providing 
a new generic description referring to this alone: " Cal. superus 5-phyllus. Cor. 
infundibuliformis; limbus planus, 5-lobus, subaequalis. Rudimentum filamenti 5-ti. 
Anther ae connexae. Caps, bilocularis." Under the name Achimenes Persoon in 
fact established a new genus distinct from Achimenes P. Browne (sensu stricto) and 
the name Achimenes has been consistently used for it. This now includes about 25 
species, of which 13 are in cultivation, together with many cultivars, some of hybrid 



i84 THE JAMAICAN SPECIES OFCOLC/MiVJE:j AHJ D A LLOPLECT U S 

origin. Being a later homonym of Achimenes P. Browne, it needs conservation; the 
type-species of Achimenes Pers. is A. coccinea (Scop.) Pers., i.e. A. ereda (Lam.) 
H. P. Fuchs. 

The type of Alloplectus 

The genus Alloplectus was founded by Martins, Nov. Gen. Sp. PI. Brasil. 3 : 53 
(1829), with the following differential character: " Calyx liber, coloratus, penta- 
phyllus, foliolis imbricatis, plus minus connatis, duobus interioribus. Corolla infera, 
tubulosa vel claviformis, rectiuscula, limbo brevi quinquelobo aut quinquedentato. 
Stamina quatuor didynama, cum quinti postici rudimento minimo, e basi tubi. 
Annulus hypogynus in glandulam posticam tumens. Capsula baccans, ovata, 
coriacea, unilocularis, bivalvis, seminibus numerosis oblongis." He included two 
Brazilian species, both described and illustrated in detail, A. sparsiflorus Mart, and 
A. circinatus Mart. 

In the list of conserved generic names in the International Code of Botanical 
Nomenclature the name Alloplectus Mart. (1829) is conserved under No. 7860 against 
Crantzia Scop. (1777) and Vireya Raf. (1814) and its lectotype given ^sA. sparsiflorus 
Mart. {= A. hirtellus (Schott) Preston), although A. circinatus would have been a 
better choice since Martins saw no mature fruits of A. sparsiflorus but had both 
flowering and fruiting material of A . circinatus and described the fruits and seeds of 
Alloplectus from this species. His two Brazilian species are, however, closely allied 
and differ in several characters from many species now referred to the genus Allop- 
lectus. If, however, a broad concept of the limits of Alloplectus is accepted, as in 
the present paper, Dalbergaria Tussac (Fl. Antilles i : 141 (1811-13) must be added 
to its synonomy as a nomen rejiciendum, since its type, D. phaenicea Tussac, loc. cit. 
(1811-13) is the same Hispaniola species [Alloplectus sanguineus (Pers.) DC, Prodr. 
7 : 546 (1839))* as Rafinesque's Vireya sanguinolenta, i.e. the type of Vireya Raf. 
already listed as a nomen rejiciendum. The short-lived Palermo Specchio delle 
Scienze Giornale enciclopedico di Sicilia,'\ wherein (i : 194 (June, 1814)) Rafinesque 
published Vireya, is an extremely scarce periodical and his account is accordingly 
reprinted here : 

" VIREYA. Cal. 5partitus, lac. equal, laciniatis, Corolla peripetala tubulosa 
5dentata, basi gibbosa; Stam.4. didynamica inclus. Ov. liberum oblong, basi 
monadeno, stigma 2lamellat. Capsula i.locul. 4valve polysperma, semina recept. 
4. longitudinal, affixa. Planta, fol. oppos. flor. axillaris. — Oss. Mi. fu comunicato 
questo nuovo genere del Sign. Turpin col. barbaro nome di Caonabo, al quale ho 
sostituito il nome dell' egregrio moderno naturalista e filosofo Virey. Si appartiene 
alia mia seconda Classe Mesogynia, 5. Ordine Epidia, Famiglia Didynamia: ha delle 
affinita colli generi Browallia, e Lindernia. 

Vireya sanguinolenta. Caule erecto simplex, foliis oppositis, alternis multo major, 
oblongis integris acutis sub-villosis subtus sanguinolentis, florib, axillarib. solitaris 
sessilis. — Oss. Nasce nellTsola di S. Domingo offerisce le singolari particolarita di 

* Martius, Nov. Gen. & Sp. 3 : 57 (1829) suggested that Besleria sanguinea Pars, belonged to Alloplectus 
but he did not make the combination A. sanguineus attributed to him by Jackson (Index Kew. i : 84 
(1893)), Urban (Symb. Antill. 2 : 358 (1901)), and Morton (Contr. U.S. Nat. Herb. 29 : 4 (1944)). 

t The British Museum (Natural History) now possesses a photostat copy. 



I 



THE JAMAICAN SPECIES OFCOL[/MiV£:^ AND^LLOPL£:crC75 185 

avere alternativamente le foglie dieci volte piii grande delle opposte, e coperte al di 
sotto di una larga macchia sanguinosa: ha le corolle d'un giallo livido." 
Vireya Raf. is thus a synonym of Dalbergaria. 

Records of Columns a and Alloplectus in Jamaica 

The first post-Linnaean author to deal with the Jamaican species of Columnea 
was Olof Swartz, who travelled extensively in Jamaica between 1784 and 1786 and 
succeeded in finding an extraordinary number of rare plants. His journeys took 
him into the parishes of Trelawny, Hanover, St. Elizabeth, St. Andrew, St. Catherine 
and St. Thomas. His collections included three new species of Columnea, i.e. 
C. hirsuta, C. hispida and C. rutilans, of which he published diagnoses in his Nova 
Genera et Species Plantarum seu Prodromus : 94 (1788) and detailed descriptions in his 
Flora Indiae Occidentalis 2 : 1080-1086 (1800). Under his C. hirsuta he cited as 
synonyms Sloane's " Rapunculus fruticosus . . ." and Browne's " Achimenes major 
herbacea . . .", but from the significant note in his description, '' Hirsuties pulchre 
articulata," and his specimen in the British Museum (Natural History) it is clear that 
he based his C. hirsuta on material of the species growing high in the Blue Mountains 
which has conspicuous multicellular hairs on the leaves. To this species the name 
C. hirsuta has always been attached. Swartz himself ascended Blue Mountain Peak 
in 1785 and, while a guest of Matthew Wallen at Coldspring, St. Andrew, in the Blue 
Mountains, he had other opportunities of collecting the species. In the west of 
Jamaica he collected C. rutilans, conspicuous for its leaves red beneath. Swartz's 
third species, C. hispida, was long enigmatic. He himself collected it only in fruit 
" in montibus altis Jamaicae occidentalis, ad rupes umbrosas ". Its flowers in fact 
remained unknown until February 1957; the flowering material collected by 
Wullschlagel in Manchester which Grisebach in 1862 (Fl. Brit. W. Ind. Is. : 465) 
referred to C. hispida belongs to another species, C. urhanii. Swartz's own descrip- 
tion and his specimens at the British Museum (Natural History) and the Rijksmuseet, 
Stockholm, show C. hispida to be a very distinct species and justify Morton's com- 
ment in 1944: " From description I am not able to identify this with any of the 
species known from Jamaica. It is perhaps a valid species, not collected again." 
Swartz, as was unfortunately his custom, gave no precise locality. In 1784 he 
travelled across western Jamaica from Montego Bay on the north coast to Savanna- 
la-Mar on the south coast and it would seem probable that he then came across this 
very rare and local species, for White Rock Hill, St. James, where Proctor and I had 
the good fortune to rediscover it in March 1956, lies near a route between these towns. 

Grisebach's account of Columnea in his Flora of the British West Indian Islands : 
464-465 (1862) added a further species, C. argentea Griseb., notable for its sericeous 
leaves and other parts, which the missionary Heinrich Rudolf Wullschlagel (1805-64) 
of the Moravian Brothers had collected in 1849 at Nazareth, Manchester. Evidently 
unaware of Grisebach's publication, Hanstein described it anew in 1865 as C. 
wullschlaegeliana. This again is a very local species. In 1956 I was overjoyed to 
find it growing among rocks near the Moravian Chapel at Nazareth, which must have 
been the very place where Wullschlagel gathered it. As noted above, Grisebach 
included under C. hispida Swartz a species here named C. urhanii. 



i86 THE JAMAICAN SPECIES OF COLUMNEA AND ALLOPLECTUS 

Grisebach placed three Jamaican species in the genus Pterygoloma, which was 
founded by Hanstein in 1854 (Linnaea 26 : 211) on a species described and iUustrated 
in Bot. Mag. 72 : t. 4250 (1846) as Alloplectus repens Hook, and there stated to have 
been collected by WilHam Purdie " in the ascent of the Sierra Nivada, Santa Martha ", 
Colombia. The first of these, which Grisebach erroneously identified with Ptery- 
goloma repens (Hook.) Hanstein and enumerated as such, is a dwarf creeping species, 
with flowers not so markedly zygomorphic as in typical species of Columnea; it had 
been collected in Jamaica both by Wullschlagel and Purdie. In 1901 Urban distin- 
guished it as Columnea jamaicensis. Grisebach's two other species, P. pubescens 
from western Jamaica and P. cristatum from central Jamaica, based on collections 
by Wilson, Purdie, Alexander and Wullschlagel, agreed with the first in their low 
creeping habit but differed in having almost cylindrical or clavate corollas with very 
short almost equal rounded lobes. They are easily distinguished from one another 
by the calyx-segments, entire in P. pubescens, pinnatifid with spreading lobes in 
P. cristatum. Fawcett listed all three under Alloplectus (see p. 234) in 1893. The 
association of the three accords with the results of a taxonometric analysis of the 
group (see p. 213). 

Nothing more was published on the Jamaican species of Columnea until 1901, 
when Urban included them in his " Enumeratio Gesneriacearum " (Symb. Antill. 
2 : 344-388 (1901)). He revised nomenclature, listed synonyms and specimens, 
and amended previous descriptions in the careful and scholarly manner which makes 
his contributions to West Indian botany so valuable as quarries of information, but 
as usual he failed to provide a key to the plants concerned, an omission which detracts 
so much from the practical utility of his work. He recorded five species as Jamaican, 
expressing his opinion about their distinctness by the comment " Fortasse omnes 
Columneae species in Jamaica obviae formis intermediis inter sese conjunctae sunt, 
C. jamaicensi excepta ". There are indeed some specimens intermediate between 
recognized species, but these would appear to result from hybridization. All the 
species accepted by Urban have definite characters associated with distinctive 
geographical ranges which justify their separation. C. hirsuta he regarded as a very 
variable species, within which, despite the limited material then available, he 
distinguished seven taxa. On the basis of the much more abundant and adequate 
material now at hand, it seems more reasonable to refer Urban' s varieties and forms 
of C. hirsuta to four species as follows: 

var. genuina (p. 361) = C. hirsuta Swartz 

var. pallescens (p. 362) = C. hirsuta Swartz 

var. concolor (p. 362) = C. hirsuta Swartz 

var. subintegra (p. 362) 

forma wullschlaegelii (p. 362) = C. urbanii Stearn 

forma hansenii (p. 363) = ? C. harrisii (Urban) Britton 

ex Morton 
forma harrisii (p. 363) = C. harrisii (Urban) Britton 

ex Morton 
var. fawcettii (p. 363) = C. fawcettii (Urban) Morton 



THE JAMAICAN SPECIES OF COLUMNEA A-^BALLOPLECTUS 187 

In 1909 Urban added another species, Columnea hrevipila, collected by Harris on 
Bluefields Mountains, Westmoreland. 

Morton in 1944 [a, b) revised the West Indian species of Columnea and Allopledus. 
With much more material for study than Urban had, most of it collected by Harris 
and by Britton, he described seven Jamaican species of Columnea in detail, raising 
two of Urban's varieties of C. hirsuta to specific rank as C. harrisii and C. fawcettii 
and adding a new species, C. subcordata, from Windsor, Trelawny. Unlike Urban, 
he provided a key to all the West Indian species of both Columnea and Allopledus. 
Examination of living material during my visit to Jamaica in 1956 (cf. Stearn, 1959) 
confirmed the taxonomic soundness of Morton's work, already evident from a prior 
survey of herbarium material. 



Definition of Columnea and Alloplectus 
The difficulties of defining the American genera of Gesneriaceae, among them 
Columnea and Alloplectus, are reflected in the different views of authors on their 
generic limits. As stated by Leeuwenberg (1958 : 293), " there are several cases of 
close inter-relationship of species usually placed in separate genera, rendering the 
boundaries between those genera vague or even untenable. Judging from these 
cases alone one would be compelled to unite large groups of genera or even complete 
tribes under one genus; this is particularly true for the Columneinae. The typical 
species of the genera are however so different that such a procedure would be quite 
undesirable. But part of those generic limits are based rather on convenient charac- 
ters than on conspicuous discontinuities of features. A consequence of this state of 
affairs is the presence of what may be called ' borderline species ', i.e. species that 
have some characters of one and some of an other genus." Such a species in Jamaica 
is Columnea jamaicensis, referred by Grisebach to Pterygoloma, by Fawcett to 
Allopledus, by Urban to Columnea. The two Brazilian species (see above) on which 
Martins founded the genus Allopledus in his Nova Genera et Species Plantarum 
Brasiliensium 3 : 53, t. 223 (1829) are markedly different from Columnea in their 
narrowly tubular rather ventricose corolla with the limb almost regular and very 
shortly lobed, i.e. shallowly 5-toothed or 5-lobed, the lobes almost equal, in their 
free anthers and their rather fleshy 2-valved capsule, but these distinctions have not 
remained clear-cut as more and more species of the group have been discovered. 
A few examples will serve to indicate the lack of correlated differences. Thus a 
species such as Columnea aureonitens Hook. (Bot. Mag. 73 : t. 4294 (1847)) has a 
narrowly tubular corolla with very short almost equal lobes, but four of the anthers 
cohere as in typical zygomorphic species of Columnea, while C. kalbreyeri Hook. f. 
(Bot. Mag. 108 : t. 6633 (1882)) has a zygomorphic corolla but the four anthers are 
completely free from one another. C. guianensis Morton, illustrated by Leeuwenberg 
(1958 • 387, fig- 16), agrees with many typical species of Columnea in having the 
leaves of a pair markedly unequal in size, but its corolla is narrowly cylindrical with 
a small almost regular limb of five short erect lobes, the two dorsal ones however 
connate, and the anthers are free from one another as in a typical Allopledus. The 
same is true of C. sanguinea (Pers.) Hanstein [Dalbergaria phaenicea Tussac) with 



i88 THE JAMAICAN SPECIES OF COLC/MATSJ AND A LLOP LE CT US 

markedly unequal leaves but a regular cylindrical corolla and anthers cohering only 
in pairs. Alloplectus savannarum Morton (illustrated by Leeuwenberg, 1958 : 365, 
fig. 9) has unequal leaves and cohering anthers but a cylindrical corolla and a two- 
valved capsule as in typical Alloplectus. A possible distinction based on the form of 
the corolla tube, i.e. gradually widened and not contracted at the mouth in Columnea, 
ventricose with a contracted mouth in Alloplectus, is obscured by such species as 
Columnea filipes Oliver (Hook., Ic. PI. 25 : t. 2428 (1896)) with ventricose tube but 
connate anthers. Characters of the fruit, an almost globose berry in typical 
Columnea, a capsule separating into two fleshy or thick valves in typical Alloplectus, 
might provide means of separation, but A. domingensis with a regular corolla has a 
berry-like fruit like a typical Columnea and the fruits of many species placed currently 
in Columnea or Alloplectus are unknown; the correlation of carpological, habit and 
floral characters in the group as a whole is accordingly doubtful. Attempts to 
separate the two genera by referring those with a coloured calyx and free anthers to 
Alloplectus, the type of which, A. hirtellus, has a dark red calyx, and those with a 
green calyx and connate anthers to Columnea are frustrated by such a species as 
C. scheideana Schlecht. (illustrated in Bot. Mag. 70 : t. 4045 (1843)) with a large 
purplish-red calyx as in Alloplectus but a markedly zygomorphic 2-lipped corolla 
and connate anthers typical of Columnea. Anisophylly is likewise unavailing, since 
the type-species of both Columnea and Alloplectus have the two leaves of a pair equal. 
In short there is no apparent discontinuity between Columnea and Alloplectus. O. 
Kuntze in 1891 (Rev. Gen. PI. 2 : 470) united the two, together with Hypocyrta and 
Nematanthus, into one genus, Columnea (sensu lato), which would now include about 
220 species, defined as follows: " Ovarium omnino superum. Disci glandula postica 
magna ceterae vel parvae. Stamina 4. Antherarum loculi paralleli. Filamenta 
basi dilatata unilateraliter connata cum corolla breviter connata. Calyx 5-partitus 
vel alte 5-fidus. Corolla apice variabilis exappendiculata. Fructus subbaccatus 
vel demum bivalvis." This course certainly has simplicity to recommend it as 
regards nomenclature but obscures the special character of Columnea (sensu stricto) 
as a highly developed group. The prevailing gamopetalous type of corolla found 
in Gesneriaceae and allied families consists of a fairly long tube with a spreading 
limb which is shorter than the tube and slightly or distinctly two-lipped; the two 
upper lobes form an upper lip while the lower lobe and the two lateral lobes together 
form a lower lip which serves as a landing platform for insects seeking nectar at the 
base of the tube. The stamens are usually included within the corolla-tube and self- 
pollination is often possible; it occurs, for example, in Alloplectus species under 
cultivation. The fruit is usually a capsule. Alloplectus (sensu stricto) represents 
this general unspecialized type. The corolla of Columnea (sensu stricto) represents 
a reversal of the trend towards development of an expanded lower lip able to support 
the weight of a bee. In Columnea the upper side of the corolla has become prolonged 
forward, the two uppermost lobes being fused and enlarged into a hood sheltering 
the exserted anthers and stigma; the two lateral lobes are reduced and they spread 
either sideways or upwards; the lower lip is also reduced and turned down or back. 
By these features, together with the horizontal poise of the flowers and their usually 
conspicuous red or yellow colouring, the flower of Columnea (sensu stricto) has 



i 



THE JAMAICAN SPECIES OF COL [/AfiV£^ AND A LLOPLECT US 189 

become perfectly adapted to pollination by nectar-seeking humming-birds on the 
wing (Text-fig. 5). The reduced lower lip and lateral lobes offer no obstacle to the 
bird as it flies under the lengthened upper lip and thrusts its beak and tongue down 
to the abundant nectar at the base of the tube, while its wings beat unimpeded on 
both sides of the lip. The anthers and stigma, being directly under the upper lip, 
are so placed that they touch the top of the head of the hovering bird. The coherence 
of the anthers ensures that they deposit the maximum of pollen in one area on the 
bird's head where the stigma of another flower will first come into contact with it. 
The corolla as a whole is firm enough in texture to sustain vibration by the wing-beats 
of a humming-bird (mostly 20-25 beats per second, but ranging from 8 to 80 according 
to the size of the bird; cf. Greenewalt, i960). Such a flower well exemplifies " the 
syndrome of ornithophily " as characterized by Faegri & van der Pijl (1966 : 109). 
This has been achieved by the successful association of several divergences from the 
almost regular flower typical of Allopledus, which divergences have evidently also 
occurred separately and independently, as the anomalous intermediate species 
mentioned above indicate. 

The proposal is accordingly made that Columnea should be defined more narrowly 
than is customary, by including within it only species of this ornithophilous habit, 
characterized by having a distinctly bilabiate red or yellow corolla, with the four 
upper (posterior) lobes fused and extended into a galea, the lower (anterior) lobe 
narrow and bent downwards, and having exserted stamens with the four anthers 
initially connate. Those species not satisfying this definition should be transferred 
to Allopledus* or possibly other genera. Thus Allopledus might likewise be more 
narrowly defined and restricted to species with a regular corolla and a capsular fruit 
akin to the type-species. To accommodate the species not fitting into this and 
Columnea (sensu stricto), Dalbergaria and possibly also Pterygoloma might then be 
restored. 

A situation of this kind, in which a well-marked specialized offshoot, such as 
Columnea (sensu stricto), is still connected to a generalized stock, such as Allopledus 

* Acceptance of this viewpoint makes it necessary to transfer to Alloplectus the following West Indian, 
Guianaian and Venezuelan species allocated to Columnea by Morton and by Leeuwenberg: 

Alloplectus affinis (Morton) Steam, comb. nov. 

Columnea affinis Morton in Fieldiana 28 : 529 (1953). 

Alloplectus aureonitens (Hook.) Stearn, comb. nov. 

Columnea aureo-nitens Hook, in Bot. Mag. 73 : t. 4294 (1847). — Leeuwenberg in Act. Bot. Neerl. 
7 : 383 (1958). 



I 



Alloplectus calotrichus (Donn. Smith) Stearn, comb. nov. 

Columnea calotricha Donn. Smith in Bot. Gaz. 40 : 9 (1905). — Leeuwenberg, tom. cit. : 385 (1958). 

Alloplectus cubensis (Urban) Stearn, comb, nov 

Columnea sanguinea var. cubensis Urban, Symb. Antill. 2 : 359 (1901). 

Columnea cubensis (Urban) Britton in Torreya 5 : 215 (1905). — -Morton in Contr. U.S. Nat. Herb. 

29 : 5 (1944)- — -Alain in Leon & Alain, Fl. Cuba 4 : 472 (1957). 

Alloplectus guianensis (Morton) Stearn, comb. nov. 

Columnea guianensis Morton in Bull. Torrey Bot. Club 75 : 564 (1948). — ^Leeuwenberg, tom. cit. : 
386 (1958)- 



igo THE JAMAICAN SPECIES OF COLUMNEA AND ALLOPLECTUS 

(sensu lato), by a few species showing intermediate stages of development, does not 
permit of a solution both logical and convenient and hence proves disconcerting to 
a tidy-minded taxonomist intent on emphasis of discontinuity, but it is all the more 
interesting and worthy of study as indicating possible lines of evolution within the 
group. The most practical procedure would seem to be to recognize as genera both 
the large heterogeneous groups manifesting a wide range of characters occurring 
in many different combinations without consistent correlations and also at the same 
time the allied apparently monophyletic groups separated from them by possessing 
consistent correlations of characters whereby these form biologically important or 
conveniently distinguishable entities, the few intermediate species being referred to 
the more heterogeneous groups. The case of Allopledus and Columnea parallels 
in some ways that of Lilium and Nomocharis (cf. Sealy, 1950) where a solution 
according with this viewpoint has proved satisfactory. 

The Jamaican species, apart from Columnea jamaicensis, fit readily into Columnea 
with a strongly bilabiate corolla and Allopledus with a regular cylindrical corolla. 
C. jamaicensis is like A. pubescens and A. grisebachianus in its low creeping habit 
and has the bilabiate character of the corolla less marked than in typical Columnea, 
but the tube expands gradually, the upper lip consists of two fused lobes with two 
spreading lateral lobes, and the four anthers are connate. These associated floral 
characters justify the allocation of C. jamaicensis, and hence of Pterygoloma, to 
Columnea, as defined here, even though on the sum of its other characters, as is shown 
by taxonometric analysis, it has more in common with the Jamaican species of 
Allopledus. 

Characters distinguishing the species 

The Jamaican species of Columnea and Allopledus are soft- wooded plants of 
creeping or sprawling habit, rather sparingly branched, with stems which may even 
reach a length of several metres in some robust specimens, the leaves then being 
clustered near the growing tip and the branchlets marked with the conspicuous scars 
of the fallen leaves. Columnea jamaicensis (PI. 14), Allopledus pubescens and 
A. grisebachianus have slender creeping branchlets only 1-3 mm. thick, whereas the 
others have much stouter branchlets about 5-10 mm. thick, initially erect but 
drooping downwards with age when the plants grow as epiphytes on trees. The 
leaves are paired, those of a pair being almost equal in size and shape, as in C. urbanii 
(PI. 16), to markedly unequal, as in C. prodorii (PI. 17) and C. rutilans (PI. 21), the 
smaller leaf of the pair having a shorter petiole as well as a shorter blade and often 
falling earlier than the larger leaf. This anisophylly, although used by Leeuwenberg 
(1958 : 295) in his study of Guiana Gesneriaceae to separate Allopledus sect. Allo- 
pledus with the leaves of a pair almost equal from Columnea sect. Collandra with 
leaves of a pair markedly unequal, varies in the Jamaican species, being apparently 
constant in some, such as C. hispida, C. subcordata and C. harrisii, but variable in 
others such as C. fawcettii and C. hirsuta, which may have the leaves of a pair almost 
equal or markedly unequal. The difference between the large and small leaves is 
particularly evident in C. rutilans, C. prodorii and C. subcordata. 



THE JAMAICAN SPECIES OFCOLf/MATK^ AND^LL0PZ.£:CT!7S 191 

The proportions as well as the actual dimension of the leaf-blade provide characters 
of some diagnostic value, as Morton demonstrated by measuring 50 representative 
leaves of each species when he had enough material available. Thus in Columnea 
jamaicensis, C. hirsuta and C. subcordata, the blades average about two times as long 
as broad, in C. fawcettii, C. brevipila and C. harrisii, between 2-4 and 2-7 times as 
long as broad, and in C. rutilans and C. argentea, about 3-4 to 37 times as long as 
broad. The leaf -base is remarkably asymmetric in C. harrisii and C. hispida 
C. rutilans stands apart from the others in having the underside of the leaf either 
completely red or else conspicuously red-veined, and this red colouring is evident 
in some of the hybrids between C. rutilans and C. urbanii. 

Diversity of hair-covering provides features of diagnostic value, notably the hair- 
covering on the upper surface of the leaf-blade, where four main types of hairs may 
be distinguished (Text-fig. i, p. 192) : 

1. Hairs erect, 8-9-celled, red, to 3-5 mm. long, e.g. Columnea hispida (Text-fig. 
lA), in which also occur minute erect colourless, 3-celled hairs. 

2. Hairs erect, 5-10-celled, colourless, o-8-2'5 mm. long, e.g. C. hirsuta (Text-fig. 
iC), C. proctorii, C. subcordata, C. urbanii (Text-fig. iB) ; in C. hirsuta these are 
associated with minute almost appressed 2-celled hairs. 

3. Hairs bent forward and somewhat or almost appressed, 4-6-celled, colourless, 
scarcely i mm. long, e.g. C. argentea (Text-fig. iF), C. harrisii (Text-fig. lE) 
and C. rutilans (Text-fig. iD), in all of which they are associated with minute 
almost appressed 2-celled hairs. 

4. Hairs bent forward, almost appressed, 2-celled, colourless, 0-2-0-4 mm. long, 
exclusively present in C. brevipila (Text-fig. iD) and C. fawcettii (Text-fig. iH), 
associated with longer hairs in C. argentea (Text-fig. iF), C. hirsuta (Text-fig. 
iC), C. harrisii (Text-fig. lE), and C. rutilans (Text-fig. iG). 

Suppression of the long hairs in C. hirsuta would produce leaves like those of 
C. fawcettii. 

The flowers are axillary, with one to five in an axil, on pedicels 1-3 cm. long, varying 
from species to species in the type of hair-covering. 

The most useful characters for specific distinction are provided by the calyx (Text- 
fig. 2). The segments may increase a little in size after flowering as the fruit matures 
but retain their shape, being, for example, about 12 mm. long in flower but 16 mm. 
long in fruit in Columnea jamaicensis, about 2-5 cm. in flower but 3-5 cm. in fruit in 
C. rutilans. They are mostly between 2 and 3 cm. long. C. subcordata (Text-fig. 
2G) stands apart from the other species, in which the five segments are free almost 
to the base, by having four segments united for about the lower third of their length 
and the fifth almost free. 

The segments are linear and entire in C. hispida (Text-fig. 2A) and lanceolate or 
narrowly lanceolate and entire or scarcely toothed in C. argentea (Text-fig. 2C), 
C. brevipila, C. harrisii, C. proctorii (Text-fig. 2B), C. urbanii (Text-fig. 2H) and 
A. pubescens (Text-fig. 29G). In C. hirsttta and C. fawcettii (Text-fig. 2D; Text-fig. 
28B) the segments bear one to three remote but usually well-marked teeth on each 
margin. In C. rutilans (Text-fig. 2F) and A. grisebachianus (Text-fig. 29C) they are 



192 



THE JAMAICAN SPECIES OF COLUMNEA AND ALLOPLECTUS 



laciniate in the lower part, with 3-5 prominent teeth on each margin. These 
differences in shape, together with differences in hair-covering, enable most species to 
be recognized by the calyx alone. The hairs of the calyx are red in C. hispida, C. 
fawcettii, C. hirsuta, C. hispida and C. rutilans, colourless in the others. 

The corolla as mentioned above (p. 186) is small, narrowly tubular and regular in 
Alloplectus grisebachianus (Text-fig. 29A, B) and A. pubescens and slightly 2-lipped 



n.n. 
3r 





D 



brevipi 



ila 



harrisii 




argentea 




rutilans 
hirsuta \ H fawcettii 




Fig. I. Hairs on upper leaf-surface of: A, Columnea hispida Swartz; B, C. urbanii Steam; 
C, C. hirsuta Swartz; D, C. brevipila Urban; E, C. harrisii (Urban) Morton; F, C. argentea 
Griseb.; G, C. rutilans Swartz; H, C. fawcettii (Urban) Morton. (All drawn from living 
plants cultivated at Kew.) 



THE JAMAICAN SPECIES OF COLUMNEA Al^D A LLOPLECT US 193 

in Columnea jamaicensis (Text-fig. 28H), but larger and strongly 2-lipped in the other 
species. It is completely yellow in C. argentea (Text-fig. 5), C. hrevipila, C. hispida 
C. subcordata and C. urhanii (Text-fig. 28G) but longitudinally striped with red in 
the others, including many of the hybrids between C. rutilans and C. urbanii. 

The filaments of the stamens are glabrous in Columnea argentea and C. brevipila 
but pubescent in most other species. 




Fig. 2. Calyx of: A, Columnea hispida Swartz {Steam 449); B, C. proctorii Steam, with 
apex of pedicel {Steam 451) ; C. argentea {Steam 359) ; D, C. fawcettii {Steam 157) ; E, C. 
jamaicensis Urban, a {Wullschldgel i2go), b {Steam 928); F, C. rutilans Swartz {Steam 
453)', G, C. subcordata Morton {Steam 472); H, C. urhanii Stearn {Steam 361). 



194 



THE JAMAICAN SPECIES OF COLC/MNEJ Al^BALLOPLECTUS 



Certain species can thus be easily recognized by distinctive and unusual characters. 
For example, Columnea jamaicensis has small leaves about 3-5 cm. long, a short 
calyx and a small corolla not strongly bilabiate. C. rutilans stands apart from 
other species in having its leaves red below and its calyx-segments partly laciniate. 
C. suhcordata has a bilabiate calyx with four segments fused for their lower third. 
C. fawcettii with toothed calyx-segments and C. hrevipila with entire calyx-segments 
differ from other species in having only very short appressed 2-celled hairs on the 
upper surface of the leaf. C. argentea has relatively narrow leaves silvery-grey with 
long somewhat appressed sericeous hairs. C. hispida is notable for the long erect 
reddish hairs covering the upper side of the leaf as well as the calyx and for the 
relatively small corolla. 

Cytology 

By Brian D. Morley 

The chromosome numbers of species of Columnea have been reported by Rogers 
(1954), Eberle (1956), Fussell (1958), Lee (1962a, 6, 1964, 1966, 1967), Morton (1963), 
Lee & Grear (1963) and Morley (1967). Investigation of the chromosome numbers 
of ten Jamaican species and four natural hybrids selected from a hybrid swarm 
between C. rutilans and C. urbanii confirms the reports for C. hrevipila, C. fawcettii 
and C. hispida by Lee (1964, 1966) and shows that all the Jamaican species have a 
diploid number of 2n = 18 and that their basic number is the same as that of other 
Columnea species, x = 9. The plants studied were as follows: 



Species Locality 

C. hirsuta Hardwar Gap, St. Andrew-Portland 

C. fawcettii Mount Diablo, St. Ann. 

C. rutilans Ramgoat Cave, Trelawny. 

C. rutilans Near Catadupa, St. James. 

C. argentea Shooters Hill, Manchester. 

C. urbanii Top Hill, Manchester. 

C. urbanii Near Newport, Manchester. 

C. urbanii Rose Hill Village, Manchester. 

C. proctorii Near Troy, Trelawny. 

C. proctorii? Ramgoat Cave, Trelawny. 

C. hispida Near Catadupa, St. James. 

C. harrisii Seven Rivers, St. James. 

C. harrisii Near Leamington, Westmoreland. 

C. hrevipila Cho Cho Gulley, Westmoreland. 

C. jamaicensis Near Leamington, Westmoreland. 



Plant 




reference no. 


2n 


Morley 33 


18 


Morley 2 


18 


(coll. E. Lodge) 




Morley 3 


18 


Morley 35 


18 


Morley 20 


18 


Morley 36 


18 


Morley 14 


18 


Morley 442 


18 


Morley S.6 


18 


Morley S.7 


18 


Morley S.17 


18 


(coll. C. D. Adams) 




Morley 5 


18 


Morley 40 


18 


Morley 23 


18 


Morley S.ii 


18 


Morley 24 


18 


Morley 27 


18 


(coll. C. D. Adams) 





THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS 



195 



C. jamaicensis 
C. rutilans X 

urbanii hybrid 4 
C. rutilans X 

urbanii hybrid 6 
C. rutilans X 

urbanii hybrid 8 
C. rutilans x 

urbanii hybrid 11 
A. grisebachianus 



Near Leamington, Westmoreland. 
Top Hill, Manchester. 

Top Hill, Manchester. 

Top Hill, Manchester. 

Top Hill, Manchester. 

Mount Diablo, St. Ann. 



Morley S.io 18 

Morley 16, 44, 45 18 



Morley 47 


18 


Morley 15, 46 


18 


Morley 48 


18 


Morley s.n. 
(coll. E. Lodge) 


18 



Although the chromosomes are small, they display sufficient morphological 
diversity to allow recognition of the following three karyotype series. A metacentric 
position is defined as an approximate position of 50 % along the length of a chromo- 
some and submetacentric as 60 %-yo % along a chromosome. 

Series i. Karyotypes with large chromosomes, i.e. longer and thicker, where the 
smallest chromosome is always longer than broad. Two pairs of chromo- 
somes with submetacentric centromere position, seven pairs of chromo- 
somes with metacentric centromere position; see Text-fig. 3. 
Species: C. hirsuta, C. fawcettii, C. urbanii (from an area north of 
Mandeville) . 

Series 2. Karyotypes with large chromosomes, i.e. longer and thicker, where the 
smallest chromosome is always longer than broad. Three pairs of 
chromosomes with submetacentric centromere position, six pairs of 
chromosomes with metacentric centromere position; see Text-fig. 3. 
Species: C. urbanii (from an area south of Mandeville), C. brevipila, C. 
rutilans, C. jamaicensis. 

Series 3. Karyotypes with small chromosomes, i.e. shorter and thinner, where the 
smallest chromosome is always as long as broad. Three pairs of chromo- 
somes with submetacentric centromere position, six pairs of chromosomes 
with metacentric centromere position ; see Text-fig. 4. 
Species: C. argentea, C. hispida, C. harrisii, C. proctorii. 

Series i karyotypes characterize the species Columnea hirsuta and C. fawcettii, 
which are morphologically and biochemically distinct from other Jamaican species 
and are only found in eastern Jamaica. Most C. urbanii karyotypes belong to 
Series 2, with the exception of one plant which had a Series i karyotype. 

Although Columnea jamaicensis is included in Series 2, it is unlike any other 
Jamaican species, having a different phenotype and a karyotype with two pairs of 
satelhtes. The metacentric satellite pair of C. urbanii (66.1460) requires further 
analysis. 

BOT. 4, 5 12 



196 THE JAMAICANSPECIES OFCOLC7MiV£'^ AND ^ LLOPLfCT C75 

Karyotypes of Jamaican species 
of Columnea . x2500 



66-991 hirsuta. 

66-1627 fawcettii. 



6610A2 



urbanii. 



»t»s* (rMltrtSissssBssss 



66-1A60 urbanii. 

\i IS as It IS tf&s8«xst 

66-960 brevipila. 

SS i( us :s M :s aa usii 

66-108A rutilans. 

66-1A68 jamaicensis. 

Fig. 3. Karyotypes of some Jamaican species of Columnea. 

The hybrids between Columnea rutilans and C. urbanii have karyotypes inters 
mediate in morphological details between those of the parents; see Text-fig. 4. 

The Jamaican species with smaller and more asymmetric karyotypes, i.e. Columnea 
argentea, C. hispida and C. harrisii in Series 3, have the more limited distributions in 
Jamaica. 

While, on evidence so far available, no changes in chromosome number have 
accompanied the evolution of Columnea, changes in chromosome structure have 
occurred, as is evident from gross morphological studies of Jamaican and Central 
American Columnea karyotypes. 



THE JAMAICAN SPECIES OF COLUMNEA Ai^B A LLOPLECT U S 197 

Karyotypes continued. 



66 ■ 1119 



argent ea. 



tt «3 ft« It s% ks )« «• i« 



66-1180 hispida 

66-1A65 harrisii. 

66-1008 subcordata. 

A ;; h ;'. r. *A n i: X I s a s « i c s « 



hybrid 3. 
66-1257 hybrid A. 

;: :a ;;& Sx :&ss :s :;s ar 

66-1291 hybrid 5. 

;;Aftss ;iir«:s«Bi's?ft« 

66-1152 hybrid 6. 

ss n3\iA Itt tx ^x %s ss ?A 

Fig. 4. Karyo-types of some Jamaican species and hybrids of Columnea. 



Chemotaxonomy 

By Brian D. Morley 

Cultivation of the Jamaican species of Columnea at Kew enabled Dr. J. B. 
Harborne to carry out on fresh leaves three chromatographic tests for flavones, 
orobanchin, and cinnamics, and also to test for saponin. The substances detected 



198 THE JAMAICAN SPECIES OF COLUMNEA K^T> A LLOPLECT U S 

in these tests were characterized as follows: 

I. Flavone a: Rf 0-35 in butanol-acetic acid-water, Rf 0-05 in 5 % acetic acid; 
colours in ultra-violet light + NH3: dark brown-bright yellow; 
possibly luteolin 7-glucoside. 
Flavone b: Rf 0-13 in 5 % acetic acid; colours as flavone a. 
Flavone c: Rf 0-23 in 5 % acetic acid; colours as flavone a. 
Flavone d: Rf 0-25 in butanol-acetic acid-water, Rf 0-05 in 5 % acetic acid; 
colours as flavone a, but more intense. 
II. Orohanchin: A complex ester of caffeic acid (see Harborne, 1966a). 
III. Cinnamics, 

Blue Spot: Rf 0-69 in butanol-acetic acid-water; colours in ultra-violet 
light : colourless-blue ; possibly a p-coumaric acid ester. 



Cinnamics, 
Turquoise 
Spot: 
IV. Saponin : 



Rf 0-24 in butanol-acetic acid-water. 

Detected by the presence of frothing when an aqueous leaf 

extract was shaken up. 



These tests yielded the following results : 









I 






II 




Ill 


IV 




Plant reference 




Flavones 




Orobanchin 


Cinnamics 


Saponin 




f 






^ 


Blue 


Turquoise 




Species 


number 


a 


b 


c 


d 




spot 


spot 




C. hirsuta 


Morley 33 


+ 


+ 


— 


— 


+ 


— 


— 


— 


C. fawcettii 


Morley 34 


+ 


+ 


— 


— 


+ 


— 


— 


— 


C. proctor a 


Morley S.17 


— 


— 


— 


— 


-t- 


+ 


+ 


+ 


C. urbanii 


Morley 12 


— 


— 


— 


— 


+ 


+ 


+ 


+ 


C. harrisii 


Morley 6 


+ 


+ 


— 


+ 


+ 


+ 


+ 


— 


C. argentea 


Morley 36 


— 


— 


tr 


— 


+ 


+ 


+ 


— 


C. hispida 


Morley 40 


— 


tr 


tr 


— 


+ 


+ 


— 


— 


C. ruHlans 


Morley 3 


— 


tr 


— 


— 


+ 


— 


— 


— 


C. brevipila 


Morley 24 


tr 


— 


— 


— 


+ 


— 


— 


— 


C. jamaicensis 


Morley 27 


+ 


— 


— 


— 


+ 


— 


+ 


— 



Where -|- = present ; — = absent ; tr = trace. 

The leaf-flavone attributes of Columnea hirsuta and C. fawcettii are similar anc 
distinct from those of the other species; they support their taxonomic separatioi 
from other Jamaican species on the basis of calyx characters. Certain cytologica 
features of the two species also distinguish them from the others in Jamaica; 
C. urbanii and C. proctorii have similar leaf-flavone attributes, and this parallels 
their morphological similarities. The leaf-flavone attributes of C. urbanii and 
C. brevipila are nearly similar, which adds some evidence for their general affinity 
to that provided by morphological and cytological data. The remaining species 
show a reticulate pattern of flavone attributes. C. harrisii was found to have a 
flavone attribute not observed in leaves of the other Jamaican species tested. 



THE JAMAICAN SPECIES OF COLUMNEA h^T> ALLOPLECT U S 



199 



The tests for cinnamics and saponin give results in accordance with those for 
flavones in the case of the species pairs Columnea hirsuta and C. fawcettii, and C. 
urhanii and C. proctorii, the first pair being negative and the second positive. The 
resemblance in flavone attributes between C. urhanii and C. hrevipila is not 
paralleled in the case of cinnamics and saponin. C. rutilans was found to have 
columnin in the leaves, a 3-desoxyanthocyanin, as reported by Harborne (19666), 
which is responsible for the red pigmentation of the abaxial leaf surface in this species. 

Columnea argentea and C. harrisii have common saponin and cinnamics attributes 
paralleling the densely hairy leaves found only in these species ; thus the sensitivity 
of the biochemical tests is shown by their discrimination of particular species within 
a group of closely related endemic species. As noted above, they provide further 
evidence of affinity between C. hirsuta and C. fawcettii, between C. urhanii and 
C. proctorii, and to a lesser extent between C. hrevipila and C. urhanii, and between 
C. argentea and C. harrisii. 

Investigation of the cinnamics and saponin attributes of material from a hybrid 
swarm at Top Hill, Manchester, gave the following results. The hybrids are arranged 
in a series suggested by their degree of morphological agreement relative to their 
parents; the biochemical attributes broadly parallel this arrangement. 







Cinnamics 




Plant 
reference 




A 




Blue 


Turquoise 


Taxon 


number 


spot 


spot 


C. rutilans 


Morley 19 


— 


— 


Hybrid 11 


Morley 48 


+ 


— 


Hybrid 10 


Morley 46 


— 


— 


Hybrid 6 


Morley 47 


+ 


— 


Hybrid 4 


Morley 8 


+ 


— 


Hybrid i 


Morley 43 


+ 


— 


C. urhanii 


Morley 31 


+ 


+ 



Saponin 



+ 



+ 
+ 
+ 



The leaf -flavone attributes of Columnea rutilans and C. urhanii are too similar to 
warrant their examination in the hybrids, but the saponin and cinnamics attributes 
of the hybrids were found to differ. The turquoise spot attribute in C. urhanii was 
not found in any of the hybrids and it may be genetically recessive. It is possible 
that backcrossing has occurred in the hybrid swarm and this may have complicated 
a simple pattern of gradation of the parental saponin and cinnamics attributes as 
found in hybrids 10 and 11. With a sufficiently comprehensive set of tests it might 
be possible to trace a backcrossing system in the hybrid swarm, particularly if 
synthetic backcrosses were made and used as marker plants. 

Harborne (19666) reported that biochemical tests distinguished between the sub- 
families Gesnerioideae and Cyrtandroideae. The present study has shown that 
biochemical tests are also sensitive at specific and intraspecific levels; leaf flavones 
provide attributes for the identification of Jamaican species of Columnea, and leaf- 
saponin and cinnamic attributes identify different hybrid phenotypes in a hybrid 
swarm. 



200 THE JAMAICAN SPECIES OF COLf/MAr£:^ K^D A LLOP LECT U S 

Harborne (19666) has suggested that there may be correlation between the posses- 
sion of 3-desoxyanthocynanins and ornithophilous polhnation in the Gesnerioideae, 
as brightly coloured flowers are attractive to birds and plants synthesizing those 
pigments in larger quantities would have a selective advantage. 

Ecology and distribution 

All the Jamaican species of Columnea are endemic. Although they do not vie in 
brilliancy of colour with such conspicuous Central American species as C. crassifolia, 
C. gloriosa, C. hirta, C. microphylla and C. oerstediana (cf. Moore, 1957; Raymond, 
1965), they are nevertheless attractive and interesting plants. They often grow as 
epiphytes high on trees and are then usually associated with bromeliads, notably 
clumps of Hohenbergia, thus easily passing unnoticed, but they also grow on rocks or 
stone walls when the rainfall is high enough to permit the luxuriant growth of mosses 
and the deposit of humus. Evidently they need good drainage and a moist or 
intermittently moist atmosphere. J 

The large zygomorphic yellow or red flowers have protruding stamens and style, ■ 
and secrete nectar at the base of the corolla-tube and are visited by humming-birds. 
Three species of humming-bird (Trochilideae) inhabit Jamaica (cf. Bond, 1947), 
namely the Jamaican Mango Humming-bird {Anthracothorax mango) with a beak 
about 2-5 cm. long, the Streamertail {Trochilus polytmus) with a beak about 2 cm. 
long and the Vervain Humming-bird [MelUsuga minima) with a beak about 9 mm. 
long. Of these, the Streamertail has been observed visiting flowers of Columnea 
(cf. Morley, 1966). The corolla-tube in most Jamaican species of Columnea averages 
about 2 cm. long; indeed the general size of the corolla appears to be correlated with 
the length of the beak and head of the Streamertail (Text-fig. 5), which has thus 
probably been a major factor in the evolution of the Jamaican species, apart from 
C. hispida and C. jamaicensis. If the Streamertail became extinct, most of the 
Jamaican species of Columnea would probably also become extinct. 

Humming-birds are strongly territorial in their habits. Thus a single bird feeds 
within a limited area and drives other humming-birds from it, thereby restricting 
pollination of ornithophilous flowers to plants within this feeding area (cf. Pitelka, 



^■•^i 




Fig. 5. Flower of Columnea argentea and head of Streamertail Humming-bird [Trochilus 

polytmus) . 



L THE JAMAICANSPECIES OFCOLf7MN£^ AND^LLOPL£:CrC75 201 

1942). Grant & Grant (1966 : 55), for example, record humming-birds each indi- 
vidually guarding its own patch of Zauschneria latifolia in flower and chasing other 
humming-birds from its preserve. These territorial habits of humming-birds 
presumably favour the inbreeding of plant colonies, which may be very important in 
the earlier stages of species development. As, however, unlike bees, humming-birds 
have no flower constancy but dart indiscriminately from one nectar-yielding species 
to another, they are likely to cross-polHnate quickly any two species flowering at the 
same time in a given feeding area. The numerous seeds contained in the white or 
pink berries of Columnea are presumably distributed from tree to tree by other birds ; 
the association of Columnea with bromeliads may possibly result from birds drinking 
water and collecting insects at the bromeliad pitchers. Ants, however, find the 
seeds attractive, according to Morley, as they do those of other Gesneriaceae, and 
thus could aid dissemination over the same tree or rocky slope. 

Although the Jamaican species have much the same ecological preferences, they 
differ markedly in range (Text-figs. 6-19) and exhibit the general tendency of the 
family Gesneriaceae towards much local endemism. Longitude yy° 20' approxi- 
mately halves the island of Jamaica. West of it occur Columnea rutilans, C. harrisii, 
C. jamaicensis, C. urhanii (with a slight extension of range east of it), C. hispida, 
C. prodorii, C. hrevipila, C. subcordata and C. argentea, to the east of it C. fawcettii 
(with a slight extension of range west of it), C. hirsuta and Allopledus pubescens, as 
also Achimenes ereda (Text-fig. 20), the early history of which is entangled with that 
of Columnea. Allopledus grisebachianus has a central position. Hybridization is 
easily effected in cultivation but rarely occurs in the wild because, although the 
ranges of some species overlap, usually only one grows in a particular locality. 
Dr. Brian Morley collected at Top Hill, Manchester, near Walderston (Text-fig. 16), 
in a man-disturbed habitat, where both C. rutilans (Text-fig. 17) and C. urbanii 
(Text-fig. 9) occur, a series of intermediate plants evidently resulting from their 
hybridization. There may also be some ecological separation. Thus C. hirsuta 
(Text-fig. 14) grows mostly between 900 m. (3 ,000 ft.) and 1,500 m. (5,000 ft.), 
although descending in Portland and St. Thomas to 600 m. (2,000 ft.). C. fawcettii 
(Text-fig. 15), which is closely allied and has been long confused with it, grows mostly 
between 400 m. (1,400 ft.) and 900 m. (3,000 ft.) in the same general area as C. hirsuta 
but extends far beyond the range of this. Some species are clearly very restricted in 
range, e.g. C. argentea (Text-fig. 12) and C. hispida (Text-fig. 7), though the reasons 
are not evident. 

Taxonometric survey 

Before the computer-aided techniques of numerical taxonomy (taxonometrics) 
can be used with confidence to classify taxa belonging to difficult or obscure groups, 
by assessing the relative over-all similarity of these taxa and then arranging them 
hierarchically in accordance with their degree of similarity and divergence, it is 
necessary to test the validity and sensitivity of such techniques by applying them to 
better understood groups already satisfactorily classified by traditional methods 
and represented by adequate material. A computer-aided technique which for such 
a group produces a classification essentially similar to or as reasonable as a well 



202 THE JAMAICAN SPECIES OF COLUMNEA A-NDALLOPLECTUS 

COLUMNEA 




jamaicensis 



hispida 





c ~~^ — . 


JAMAICA 


y^-.r ■ 


H : . 


' — " — ^— w^^ 




^■^ — s ,■■■ ■■ * 




^■k-K 


JU.t,, 


VM 


^ 'r^ 


'j^'^^t^ 




^i 














8 proctorii 



urbanii 





10 subcordata 



1 1 harrisii 



1 J 


JAMAICA 


V'; ■■; '■'•■ • 


#, ' \ ! .■ ''■■■...:.^\ 


. . ^■>:: .. .. 


\\ 


■\. w/^^^*~Cj._.~^ 




"^6 ! ; 






12 argentea 



1 3 brevipila 



Figs. 6-13. Distribution of Columnea in Jamaica: Fig. 6, C. jamaicensis; Fig. 7, C. hispida; 
Fig. 8, C. proctorii; Fig. 9, C. urbanii; Fig. 10, C. subcordata; Fig. 11, C. harrisii; Fig. 12, 
C. argentea; Fig. 13, C. brevipila. 



THE JAMAICANSPECIES OF COLUMNEA At^BALLOPLECTUS 203 

COLUMNEA 





14 hirsuta 



15 fawcettii 



JAMAICA 




JAMAICA 




16 hybrid (rutilans x urbanii) 17 

Alloplectus 



rutilans 



JAMAICA 



"4.* 



- — \-.^_,I_ '" " 1'"' 

'. I \ 



Uf,-!_;^i_i» 



JAMAICA 




grisebachianus 19 pubescens 

ACHIMENFS 



J 


,«■ ,. 77- „■ 1 


.1 




JAMAICA 


'\ -1 ^ j^^. 


M,l„ 








,f ,. - ,.• 1 



20 



erecta 



Figs. 14-20. Distribution of Columnea, Alloplectus and Achimenes in Jamaica: Fig. 14, 
Columnea hirsuta; Fig. 15, C. fawcettii; Fig. 16, C. rutilans x urbanii; Fig. 17, C. rutilans; 
Fig. 18, Alloplectus grisebachianus; Fig. 19, ^4. pubescens; Fig. 20, Achimenes erecta. 



204 THE JAMAICAN SPECIES OF COLC/MATS^ AND A LLOP LECT US 

grounded existing one obviously commends itself for use in other investigations. 
Conversely a technique which makes nonsense when dealing with the known is 
unlikely to do better with the unknown. Application of taxonometric methods to 
different groups has given results of varied acceptability. Thus a comparison by 
Kendrick and Weresub (1966) of members of orders within the Basidiomycetes, 
using 67 haphazardly assembled equally weighted characters, resulted in a haphazard 
classification, seemingly of no value for any purpose. Watson, Williams & Lance 
(1967), using 38 characters, produced a classification of the genera of Ericales which 
bears only a limited resemblance to the one generally accepted ; they regard, however, 
the new scheme as more convincing, " since it provides a convenient summary of the 
information put into it and makes sense of observations other than those on which 
it was based, while the traditional scheme does neither." Sheal's taxonometric 
survey of mesostigmatid mites of the Hypoaspis-Androlaekps complex, using 62 
characters, gave results in good agreement with a recent revision by traditional 
methods (Sheals, 1965; cf. also British Museum (Natural History), 1966). Likewise 
an arrangement of the genera of megascolecoid earthworms (Oligochaeta Megas- 
colecidae) by computer technique, using 43 characters, gave results coinciding largely 
with one of the previously proposed classifications (Sims, 1966). 

A major difficulty in numerical taxonomy is finding enough attributes for computa- 
tion; obviously the smaller the number, the greater the risk of results distorted by 
abnormalities. Between 40 and 100 characters have been postulated as necessary 
to give results statistically valid (Sokal & Sneath, 1963), but in practice it is often 
difficult or impossible to provide so many characters; the diversity within small 
groups does not stretch that far. Hence it is desirable to ascertain empirically 
whether a much smaller number of characters will yield results which, even if not so 
detailed and meaningful as those derived from a large number of characters, can 
nevertheless be helpful by grouping together organisms with over-all resemblance 
and separating those which differ. It is also desirable to ascertain the extent to 
which a reduced set of selected characters and an equally reduced set of characters 
picked at random agree in their results with each other and with a maximum set. 

The information available about the Jamaican species of Columnea and Allopledus 
seemed to be broadly enough based and adequate enough in quantity to provide such 
comparisons and to test the ability of a computer-aided taxonometric procedure to 
classify these species. By including cytological, biochemical and anatomical data 
along with gross morphological data it was possible to tabulate 51 characters varying 
from species to species, i.e. constant within a given species but not for the whole 
group. This full set of characters was the basis of one computation; 27 of them had 
been used in a key and thus formed a reduced set weighted in favour of readily 
observed macroscopic characters, on which indeed most angiosperm taxonomy is 
necessarily based. These were used for a second computation. For comparison, 
27 characters were taken at random, i.e. by assigning number to each of the full 51 
characters and picking blindfold 27 cards out of 51 bearing these numbers; 13 were 
found to coincide with those of the other set. These were used for a third computa- 
tion. Such characters for coding purposes may be divided into three groups (cf. 
Watson, Williams & Lance, 1967) : qualitative (or discrete) two-state (or dichotomous), 



I 



THE JAMAICAN SPECIES OF COLUMNEA AND ALLOPLECTUS 205 

in which a character exists in one or other of two states; qualitative multi-state, in 
which a character exists in one of three or more states; quantitative (or continuous) 
comprising characters within which several measurable or assessable though less 
clear-cut and sometimes overlapping states may be distinguished. For some 
characters their allocation to one or other of these groups is arbitrary. Those used 
in this survey were as follows : 



TWO-STATE 

I.* Stem thickness: less than 5 mm.; more than 5 mm. 

2.* Leaves: isophyllous; anisophyllous. 

3.* Leaf base: symmetric; asymmetric. 

4.* Leaf underside: green; red. 

5.* Hairs of leaf: colourless; red. 

6. 2~j-celled hairs on leaf upperside: absent; present. 

7. Other hairs on leaf upperside : 4-7-celled; 8-io-celled. 
8.* 2-3-celled hairs on leaf underside: absent; present. 

9. Other hairs on leaf underside: 4-7 celled; 8-io-celled. 

10. Pedicels: always i; i-several. 

11. Hairs of pedicels: colourless; red. 

12.* Calyx-segments: all free; 4 fused, r free, 

13.* Calyx: green; wholly or partly red. 

14.* Hairs of calyx: long; short. 

15. Long hairs of calyx: erect; otherwise. 

16. Short hairs of calyx: few; many. 
17.* Hairs of calyx: colourless; red. 

18.* Corolla: completely yellow; yellow and red. 

19.* Corolla length: 3 cm. or less; 4 cm. or more. 

20.* Corolla-tube length: 1 cm.; I-4-2-5 cm. 

21. External hairs of corolla: 2-6-celled; 6-12-celled. 

22. External hairs of corolla: colourless; red. 

23. External hairs of corolla: appressed; spreading. 

24. External hairs of corolla : few ; many. 

25. Internal hairs of corolla: 1-3-celled; 3-6-celled. 
26.* Filaments of stamens: glabrous; hairy. 

27. Non-glandular hairs on style: absent; present. 

28. Submetacentric chromosome pairs: 2; 3. 

29. Saponin: absent; present. 
30.* Corolla: regular; zygomorphic. 
31. Anthers: almost square; oblong. 



MULTI-STATE 

32. Flavone: none; A; B; C. 



2o6 THE JAMAICAN SPECIES OF COLUMNEA AT^DALLOPLECTUS 



33 

34 
35 
36 
37 

38 
39 

40 

41 
42 

43 
44 
45 
46 

47 
48 

49 
50 
51 



QUANTITATIVE 

Leaf-blade length I breadth: 1-5-2; 2-4-2-7; 3-4-4. 

Leaf-blade length: under 5-5 cm.; 5-5-12 cm.; more than 12 cm. 

Length of long petiole : almost nil; to i cm.; to 1-5 cm.; to 3-5 cm. 

Vein pairs of leaf : 3; 4-5; 5-7; 7-8. 

Hair density on leaf upperside per sq. cm.: 0-25; 26-100; 101-400; 401-800; 

801-1,600. 
Hairs on leaf upperside : erect; strongly curved; appressed. 
Hair density on leaf underside per sq. cm. : 0-25 ; 26-100 ; 101-400 ; 401-800 ; 

801-1,600. 
Hairs on leaf underside : erect ; strongly curved ; appressed. 
Palisade layer: in upper |; at middle; in lower ^. 
Cell width in palisade layer: narrow; medium; broad. 
Pedicel length: 2-10 mm.; to 25 mm.; to 30 mm. 
Hairs on pedicel: appressed; ascending; erect. 
Length of calyx-segments: 1-1-4 cm.; 1-5-1-9 cm.; 2-2-4 cm- 
Calyx-segments : entire or almost so ; toothed ; laciniate. 
Length of upper lip of corolla: nil; 2-2-4 cm.; 2-5-3 cm.; 3-I-3-4 cm. 
Length of mid lobe of corolla: nil; 3-5 mm.; 10-15 mm.; 16-18 mm. 
Length of lower lip of corolla: nil; i cm. or less; 1-5 cm. or more. 
Ovary : glabrous ; hairy above ; completely hairy. 
Cinnamic: absent; scanty; abundant. 



These characters were ascertained from living and herbarium material and codec 
numerically, e.g. hairs colourless = 0, hairs red = i, hairs appressed — 0, hairs 
strongly curved = i, hairs erect = 2, then tabulated as a species/data matrix, all 
being included in the full matrix, but only the 27 marked * in the reduced matrij 
of selected characters (reduced key matrix). For many quantitative characters, an 
average from a number of specimens was taken and the coding based on this. 
Computation of the above data on an Orion electronic computer at the Rothamsted 
Experimental Station, Harpenden, Herts., England, by G. Ross, using Gower & 
Ross's program CLASP, gave the " median sort " results expressed as dendrograms 
(phenograms) in Text-figs. 21-23. 

Since certain species are obviously closely linked in their morphological characters, 
e.g. (i) Columnea hirsuta and C. fawcettii, (2) C. subcordata and C. proctorii, (3) d 
argentea, C. brevipila and C. harrisii, and others are relatively isolated, e.g. C. 
hispida, C. jamaicensis, A. grisebachianus , it was postulated before computation that, 
unless the resulting dendrograms linked and separated such species accordingly, 
they would have little or no relevance to the actual situation within the genus. In 
fact all three dendrograms make these associations and distinctions. The dendro- 
gram (Text-fig. 21) from the full matrix (51 characters) achieves, however, the most 
satisfactory differentiation, then comes that from the reduced random matrix 
(Text-fig. 22) , both being superior to that from the reduced key matrix of 27 charac- 
ters (Text-fig. 23). Thus, whereas the dendrogram from the reduced key matrix 



THE JAMAICAN SPECIES OF COLUMNEA ANT) A LLOPLECT US 



207 



(Text-fig. 23) simply places together at a similarity coefficient (phenon level) of 80 % 
C. fawcettii, C. harrisii, C. hirsuta, C. proctorii and C. suhcordata without further 
distinction and separates from them at the 75 % level C. argentea and C. brevipila, 
the full one and the reduced random one put these species in three groups, the full 
one (Text-fig. 21) at the 75% level, the random (Text-fig. 22) at the 70% level, i.e. 
(i) C. hirsuta and C. fawcettii together, (2) C. subcordata and C. proctorii with C. 
urbanii, (3) C. argentea and C. brevipila with C. harrisii; the dendrogram from the 
full matrix goes further by associating C. hirsuta, C. fawcettii and C. rutilans. In 



65 



70 



75 



80 



85 



90 



95 



_^.^___ ambiguus 

— i— ^— i— pubescens 
__i^_^__^ grisebachianus 

-_i._^.._ hispida 

I argentea 

———-^^ brevipila 

— ^_i— _ harrisii 

I fawcettii 

__^.._i^ hirsuta 

— ^-^-^^ hybrid 6 

^— ^^^_ rutilans 

^____ hybrid 10 

— .i...i— _ hybrid 4 

^^^__ urbanii 

-i__— ^^ proctorii 

--____^_ subcordata 

— ^— — ^^— — — — '^^ ^— ^— — _^_ii...^..^^^_^_ jamaicensis 

Fig. 21. Dendrogram showing the percentage of similarity of 17 species and hybrids of 
Columnea and Alloplectus computed from a matrix of 51 characters (cf. Fig. 24). 



208 



THE JAMAICAN SPECIES OY COLUMNEA ANDALLOPLECTUS 



these instances, a taxonometric method has thus created from an assemblage of 
numerous unweighted characters a classification basically the same as one made by 
traditional methods. All dendrograms emphasize the distinctness of C. jamaicensis 
from other Jamaican species of Columnea by separating it at the 65% (full and 
reduced random matrices) and 70% (reduced key matrix) levels; no dendrogram, 
however, indicates any resemblance of this to the Alloplectus species A. ambiguus, 
A. grisebachianus and A. pubescens. All associate the hybrids with one or other of 
their parents. The superiority of the dendrogram based on 27 random characters 
(Text-fig. 22) over that based on 27 selected characters (Text-fig. 23) would appear 
to result from the first including biochemical, anatomical and cytological characters 
as opposed to the purely morphological characters of the second. The difference 
between them is in line with the general view of taxonomists that certain characters 



65 



70 



75 



80 



85 



90 



95 



100 



^ii— — "^ ambiguus 

"""^"■"^ pubescens 

•"— ~~~~~ grisebachianus 

— ^^^^^^ harrisii 

— — ■^— ■ argentea 

^^^— — ■ brevipila 

^^— — iawcettii 

^-^-^-i— hirsuta 

— — ^— ^ hybrid 6 

^•^^m^^^ hybrid lo 

^_^^— H rutilans 

I hybrid 4 

__i^^_a urbanii 

■ subcordata 

__^..^ proctorii 

^_^^_^ jamaicensis 

^____^ hispida 

Fig. 22. Dendrogram showing the percentage of similarity of 17 species and hybrids of 
Columnea and Alloplectus computed from a matrix of 27 characters taken at random (cf. 
Fig. 25). 



L 



THE JAMAICAN SPECIES OF COLUMNEA A'^T) A LLOPLECT U S 



209 



reflect more strongly than others the genetic composition of taxa but it also indicates 
that such characters are not all of them consciously used by taxonomists. 

A dendrogram provides a convenient expression of relationship by degree of 
similarity, since this similarity presumably arises from and expresses a similarity of 
genes, but it suffers from the disadvantages inevitable in a linear sequence. A more 
revealing view of relationships may be obtained by producing from the similarity 
matrix, by means of Gower's method of Principal Co-ordinates Analysis (Gower, 



I 



65 



70 



75 



80 



85 



90 



95 



— ._^_ ambiguus 

I pubescens 

_^.^___„ grisebachianus 

^^_..i.^ argentea 

^— ^^^._ brevipila 
hybrid 4 

^— — ^— urbanii 

_^_^__« fawcettii 

^__^.i_ harrisii 

____ hirsuta 

^__._ proctorii 

_ ^^— — subcordata 

• hybrid 6 

■ hybrid 10 
^_^^-. rutilans 
——^ jamaicensis 

■ hispida 

Fig. 23. Dendrogram showing the percentage of similarity of 17 species and hybrids of 

Columnea and Alloplectus computed from a matrix of 27 characters used in a key for their 
recognition (cf. Fig. 26). 



2IO THE JAMAICAN SPECIES 0¥ COLUMNEA ANBALLOPLECTUS 



5,. 



rutilans 



hybrid 6 



COLUMNEA 



hybrid lo 
hirsuta • 9 

fawcettii 



-4 -3 -2 -I 

I I I I I 



hispida 



brevipila 

, . .. • proctorii 
narnsii m , , ^ 

• subcordata 

• 
• • argentea 
hybrid 4 

I 2 3 

I I » » » 



urbanii 



-2 



Jamaicensis 



ambigi 



irisebachianus 



pubescens 



Alloplectus 



Fig. 24. Vector diagram showing relative similarity of 17 species and hybrids of Columnea 
and Alloplectus computed from a matrix of 51 characters (cf. Fig. 21); the vertical 
co-ordinates are the vectors corresponding to the first latent roots, the horizontal co- 
ordinates those of the second latent roots, of the transformed matrix. 



THE JAMAICAN SPECIES OF COLUMNEA Al^D A LLOPLECT US 211 



COLUMNFA 



brevipila 



5 
4 
3 
2 



rutila 



hybrid 6 



hirsuta 
hybrid 10 • 

fawcetti 



proctorii 

C hybrid 4 



■', -^ -' 



harrisii • * subcordata 

• argentea 

urbanii 



-2 



hispida 



jamaicensis 



grisebachianus • 



I ambiguus 

pubescens ° 



Alloplfctus 



Fig. 25. Vector diagram showing relative similarity of 17 species and hybrids of Columnea 
and Alloplectus computed from a matrix of 27 characters taken at random (cf. Fig. 22); 
co-ordinates as in Fig. 24. 

BOX. 4, 5. 13 



212 THE JAMAICANSPECIES OFCOL[7MAr£:j K^T> ALLOPLECT U S 

1966), a scatter diagram with the scaled first and second vectors of the transformed 
matrix as co-ordinate axes. In this vector diagram the distances between the units 
are proportional to the square root of their dissimilarity. Those resulting from the 
full and reduced matrices (Text-figs. 24-26) differ very much less than might be 
expected; indeed the major differences between the three are in the relation of the 
units to the axes and not to one another, as becomes evident by rotating the diagram 
of the reduced key matrix (Text-fig. 26) through about 45 degrees to the left and by 



COLUMNEA 




Alloplfxtus 



pubescens • 



Fig. 26. Vector diagram showing relative similarity of 1 7 species and hybrids of Co/wmw^a ]A 
and A lloplectus computed from a matrix of 27 characters used in a key for their recognition ^^ 
(cf . Fig. 23) ; the right to left co-ordinates are derived from the first latent roots, the left 
to right co-ordinates from the second latent roots (cf. Fig. 24). 



THEJAMAICANSPECIES0FC0Li7MAr£^ AND ^ LLOPLECT t/S 213 

transposing from left to right and vice versa the units of the diagram of the reduced 
random matrix (Text-fig. 25). They all convey much the same general information. 
In all three diagrams, Columnea hirsuta and C. fawcettii are brought closer to C. 
rutilans than is apparent in the dendrogram from the reduced matrices; C. ruHlans 
and C. urhanii are well isolated and their hybrids spaced between them. All 
diagrams place fairly close together C. proctorii, C. brevipila, C. subcordata, C. 
argentea and hybrid 4, an assemblage which should, however, on morphological 
grounds be separated into three groups. The configuration of species becomes here 
more meaningful if the third latent vector is brought into consideration by being 
treated as a third dimension. Thus, in the data derived from the full matrix, the 
third latent vector of — 0-25 for C. proctorii brings this close to C. subcordata with a 
vector of — 0-24, while C. brevipila, with a vector of 0-25, and C. argentea, with a 
vector of 0-23, stand apart from them both and from hybrid 4, with a vector of o-ii ; 
consideration of the third latent vector thus separates them into three groups agreeing 
with their general morphological characters. 

The units on the vector diagrams resolve into a general pattern which appears 
significant. The typical species of Columnea with large zygomorphic flowers stretch 
together across the upper part of the diagram of the full matrix (Text-fig. 24) in a 
sequence passing from C. urbanii with green isophyllous leaves and entire calyx- 
segments to C. rutilans with markedly anisophyllous leaves red below and laciniate 
calyx-segments, and they undoubtedly form a natural group, the evolution of which 
has probably been partly controlled by the Jamaican endemic Streamertail Humming- 
bird (Text-fig. 5). From this group C. hispida stands somewhat apart, and in view 
of its short and more dingy corolla it probably has a different pollinating agent. Far 
belov' them and well separated from each other in every diagram are the species of 
Allopiectus, i.e. A. ambiguus, A. grisebachianus and A. pubescens, probably self- 
pollinated, with C. jamaicensis brought much closer to them than to the typical 
species of Columnea, a placing which agrees well with its general morphology but is 
not evident from the dendrograms. 

The vector diagrams give indeed the general impression that, whereas the typical 
Columnea species of Jamaica can well be regarded as derivatives of a single stock, 
C. jamaicensis represents an independent and florally less developed offshoot of 
Allopiectus. 

A dendrogram simply indicates degree of similarity without necessarily placing 
the units of a particular group into a meaningful sequence within it, whereas a scatter 
diagram places them in two-dimensional relation to one another and hence can be 
used as a better basis for arranging the units in a sequence indicative of resemblance 
and possibly of genetic affinity. The dendrogram derived from the full matrix 
arranged in this way (Text-fig. 27) by reference to the corresponding scatter diagram 
(Text-fig. 24) gives an arrangement which coincides almost exactly with the taxo- 
nomic arrangement previously made by traditional methods. 

This survey thus confirms the view that taxonometric methods when applied to an 
adequate body of data can produce a grouping of species certainly as satisfactory as 
one obtained by traditional methods and possibly more suggestive of their course of 
development; moreover such groupings can result from the computation of much 



214 



THE JAMAICAN SPECIES OF COLUMNEA Al^BALLOPLECTUS 



less than the postulated necessary 40 characters, particularly when the characters 
used are diverse enough to epitomise the whole organization of the species. 

SYSTEMATIC ACCOUNT 

Material studied 

This account has been based on the collection in the Department of Botany, 
British Museum (Natural History), London (BM) together with material lent by the 

65 70 75 80 85 90 95 



1 



rutilans 

hybrid 6 

hybrid 10 

hirsuta 

fawcettii 

brevipila 

argentea 

harrisii 

proctorii 

subcordata 




)ubescens 



jamaicensis 



Fig. 27. Dendrogram showing phenetic affinity of 17 species and hybrids of Columnea 
and Alloplectus computed from a matrix of 51 characters and arranged in a sequence 
based on a vector diagram from the first and second latent roots (cf. Figs. 21, 24). 



THE JAMAICAN SPECIES OF COLUMNEA AND A LLOPLECT US 215 

Systematisch-Geobotanisches Institut, Universitat, Gottingen (GOET), the Science 
Museum, Institute of Jamaica, Kingston, Jamaica (IJ), the Royal Botanic Gardens, 
Kew (K), the Botanische Staatssammlung, Munich (M), the New York Botanical 
Garden (NY), the Department of Botany, University of the West Indies, Mona, 
Jamaica (UCWI) and the United States National Herbarium, Smithsonian Institu- 
tion, Washington, D.C. (US). 

Specimens are listed by the parishes of Jamaica, beginning with Hanover and 
Westmoreland in the west and ending with Portland and St. Thomas in the east 
(cf. Stearn, 1959 : 134, fig. la). 

As an indication of how our knowledge of the Jamaican species of Columnea and 
Alloplectus has grown, it is of interest to note that Grisebach (in 1862) distinguished 
seven species on 15 gatherings. Urban (in 1901) seven species on 20 gatherings, 
Morton (in 1944) ten species on 54 gatherings; the present account distinguishing 13 
species has been based on 190 gatherings. Our understanding of the group in 
Jamaica would appear more likely now to benefit from field observations on pollina- 
tion and dissemination than from further collecting. 



Acknowledgements 

Grateful thanks are here expressed to those in charge of the above herbaria for the 
opportunity to study their material. I am indebted especially to Mr. George R. 
Proctor for his help to me while collecting Columneas and other plants in Jamaica. 
Dr. Brian D. Morley has generously made available the results of his extensive 
cytological and other research and contributed the sections on cytology and 
biochemistry above. He and Miss Barbara Heywood also assisted me in ascertaining 
and tabulating the characters used in the taxonometric analysis above, which is the 
result of computation by Mr. Gavin J. S. Ross at the Department of Statistics, 
Rothamsted Experimental Station ; it has benefited much from the advice of Dr. J. G 
Sheals. Mr. D. Erasmus drew the text-figures. The co-operation of them all is 
much appreciated. 



Key to Jamaican species of Columnea and Alloplectus 

Corolla tubular or narrowly funnel-shaped, the limb not markedly 2-lipped, the 

lobes of the limb only 2-8 mm. long; leaf -bearing branches slender (1-3 mm. 

diam.) ; all leaf-blades small (mostly under 5 cm. long, 3 cm. broad) : 

Calyx-segments laciniate with four or five prominent teeth either side; central 

Jamaica . . . . . . 2. Alloplectus grisebachianus 

Calyx-segments entire or with one or two minute teeth either side : 

Leaves and calyx densely hirsute; corolla-lobes almost uniform, about 1-2 mm. 
long ; eastern Jamaica . . . . . . . 1. A. pubescens 

Leaves and calyx apparently glabrous but with ciliate margins and short 
inconspicuous hairs on surface; corolla-lobes forming two lips, the lower 
lip about 6 mm. long; western Jamaica . . i. Columnea jamaicensis 



2i6 THE JAMAICANSPECIESOFC0LI[7MN£^ A-!<iT> A LLOPLECT US 

Corolla markedly 2-lipped, the lobes of the limb i cm. or more long; leaf-bearing 
branches stout (5-10 mm. diam.) ; leaf-blade of at least one leaf of a pair larger 
(longer than 5 cm. or broader than 3 cm.) : 
Leaf-blades partly or completely red beneath or coloured red along the veins: 
Calyx-segments deeply toothed along the margin, the lowermost teeth to 5 mm, 
long; larger leaf -blades three to four times as long as broad 11. C. rutilans 
Calyx-segments entire or shortly toothed, the teeth to 2 mm. long; larger leaf- 
blades about two times as long as broad . . C. rutilans x urhanii 
Leaf -blades green or silvery beneath, not red- veined or red: 

Hairs on upper side of leaf minute, 0-2-0-3 mm. long, 2-celled, inconspicuous: 

Calyx-segments entire, green, inconspicuously hairy with short rather 

appressed hairs : 

Pedicels with minute appressed hairs 0-4 mm. or less long; Westmoreland, 

St. Elizabeth . . . . . . . 8. C. brevipila 

Pedicels with longer somewhat spreading hairs to i mm. long; Manchester 

C. rutilans X urhanii 

Calyx-segments with one to three linear teeth to 2 mm. long on each margin, 

red in upper part, densely hirsute . . . . 9. C. fawcettii 

Hairs on upper side of leaf i mm. or more long, 3-8-celled, conspicuous: 
Calyx sericeous with dense appressed silvery hairs : 

Leaf -blades densely sericeous beneath, three to five times as long as broad, 
with four or five primary veins either side, almost symmetric at base; 
corolla yellow . . . . . . . 6. C. argentea 

Leaf-blades with short curved hairs not completely covering the surface 
beneath, about two to three times as long as broad, with six to eight 
primary veins either side, markedly asymmetric at base; corolla red 
and yellow . . . . . . . . 6. C. harrisii 

Calyx not sericeous but hispid or pilose with long, spreading, often glandular, 

hairs : 

Calyx-segments with one to three linear teeth on each margin and with 

attentuate red tips; eastern Jamaica . . . 9. C. hirsuta 

Calyx-segments entire or only minutely toothed, green, the tips not 

attenuate ; central and western Jamaica : 

Leaves and calyx with conspicuous red hairs; corolla about 2-5 cm. long 

2. C. hispida 

Leaves and calyx with colourless (not red) hairs; corolla 4-5 cm. long: 

Four calyx-segments joined at base for 5 to 8 mm., the fifth (posterior) 

one free from them; leaf-blades sometimes subcordate at base, 

usually asymmetric . . . . . ■ 5- C. subcordata 

All calyx-segments free almost to the base; leaf-blades rounded to 

cuneate at base: 

Leaves of a pair markedly unequal in size, the largest asymmetric at 

base and usually 8-14 cm. long, hirsute below with abundant 

long spreading hairs ; corolla striped red and yellow throughout 

4. C. prociorii 



t 



THE JAMAICAN SPECIES OF COLC/MATS^ AND A LLOPLECT U S 217 

Leaves of a pair almost equal in size, the largest symmetric at base 

and at most 8 cm. long, usually much less, strigillose below 

with very short appressed hairs; corolla yellow or with red 

markings at base only: 

Corolla yellow ; style yellow .... ■^.C.urhanii 

Corolla yellow with red stripes on base, or, if entirely yellow, style 

red . . . . . . . C. rutilans x urhanii 

COLUMNEA L. 

CoLUMNEA L., Sp. PI. 2 : 638 (1753) ; Gen. PI. ed. 5 : 286 (1754). 

Shrubs. Leaves opposite, those of a pair equal to markedly unequal, usually 
shortly petiolate. Flowers solitary or fasciculate in leaf-axils, with or without 
bracts. Calyx 5 -partite. Corolla inferior, erect in the calyx, gibbous but not spurred 
at base, the tube gradually expanded upwards, long, the limb 2-lipped, the upper lip 
3-lobed with the middle lobe formed from two united lobes into a notched galea or 
hood and with two lateral lobes spreading from it, the lower lip entire and recurving 
or reflexed. Stamens 4, didynamous, the filaments united in the lower part, free 
and filamentous above, the anthers of all four stamens cohering at anthesis, their 
loculi parallel, longitudinally dehiscent; fifth stamen represented by a staminode. 
Disc gland large and dorsal, sometimes accompanied by two to four smaller glands. 
Ovary superior; stigma usually 2-lobed. Fruit a fleshy berry; seeds minute, longi- 
tudinally or obliquely striate. 
Type-species by monotypy: C. scandens L. 

The above description does not cover the anomalous non-Jamaican species with 
regular corolla mentioned above (p. 187). 

Columnea jamaicensis Urban, Symb. Antill. 2 : 359 (1901). — Morton in 
Contr. U.S. Nat. Herb. 29 : 3 (1944). — Stearn in Proc. Linn. Soc. Lond. 170 : 141 
(1959). (PI. 14; Text-figs. 2E, 6, 28H.) 

Pterygoloma repens sensu Griseb., Fl. Brit. W. Ind. Is. : 464 (1862) pro parte, quoad pi. 
Jamaic. ; non Hanstein. 

Shrub of dwarf creeping habit, the stems grey, slender, c. 2 mm. thick, pilose with 
reddish hairs when young, later glabrous. Leaves paired, those of a pair unequal, 
sparsely setose on both surfaces, dark green above, paler green below, the larger leaf 
of the pair long-stalked, the blade narrowly ovate, asymmetric, with the apex shortly 
acuminate or acute, the margin almost entire or remotely serrulate, the base cuneate, 
2'5-5*5 cm. long, 1-2-3 cm. broad, the smaller leaf soon falling, short- or long-stalked, 
the blade 0-5-3 cm. long, 0-5-1-5 cm. broad; petiole of larger leaves 5-1 1 mm. long, 
of smaller leaves 1-3 mm. long. Flowers solitary in leaf-axils. Pedicel 1-5-3 cm. 
long, strigillose with very short appressed hairs. Calyx divided to the base, reddish 
or green, sparsely and shortly pilose and ciliate outside; segments narrowly ovate or 
lanceolate, acute, 8-5-12 mm. long in flower, to 16 mm. long in fruit, entire or with 
one or two minute teeth on each margin. Corolla tubular, gradually widened, 2-5- 



2l8 



THE JAMAICAN SPECIES OF COLUMNEA A-NBALLOPLECTUS 



3 cm. long, crimson or orange yellow streaked with red bands, the tube much longer 
than the calyx, the upper lip about 0-5-1 cm. long, 3-lobed, the lower lip 5-8 mm. 
long, narrowly oblong. Stamens slightly exserted; filaments glabrous; anthers 
about I mm. long. Ovary glabrous except for a few sparse hairs near the apex; style 
glandular-pilose. Berry white, c. 5 mm. across. 

Westmoreland: Springfield, 1849, Wullschldgel i2go (M; GOET, type of C. 
jamaicensis). Bluefields Mountain, four miles from the sea, 600 m., Purdie (K). 




Fig. 28. Columnea fawcettii (Urban) Morton: A, flowering shoot; B, calyx-segment; C, 
stamens; D, anthers; E, anther showing insertion of filament; F, gynoecium, with 
disc-gland [Steam 255) . Columnea urhanii Stearn ; G, flower [Steam 361, spirit material) . 
Columnea jamaicensis Urban: H, flower [Steam g28, spirit material). 



THE JAMAICAN SPECIES OFCOLC7MiV£:j AISST) A LLOPLECT US 219 

Three miles west of New Works, Bluefields Hill, 680 m., July 1954, Farr 5643 (IJ). 
Two and a half miles northeast of Bluefields P.O., 700 m., July 1954, Webster & 
Wilson ig54 (AA). Kentucky Hill, Bluefields Mountain, 600-760 m., March 1908, 
Harris 10200 (BM; NY). Woodstock, near Newmarket, 420 m., Sept. 1907, Britton 
1 571 (NY; US), Harris 9836 (NY). Same locality, Oct. i960. Proctor 21 518 (BM). 
Near Woodstock, 360 m., Dec. 1965, Adams I2yg4 (BM). One mile west-north- 
west of Hopewell, 530 m., Nov. 1955, Proctor 11232 (BM; IJ). 

Trelawny: Cockpit Country, at Black Banana Hole, five miles north of Accom- 
pong, 330 m., April, 1956, Steam 928 (BM). Red Mud Hole, six miles north-north- 
west of Accompong, 330 m., April 1956, Steam 964 (BM). 

Urban based his C. jamaicensis on specimens collected by Wullschlagel at 
" Springfield ", presumably Springfield in Westmoreland parish, although there 
are at least 20 other places in Jamaica named Springfield. Grisebach had earlier 
(Fl. Brit. West. Ind. Is. : 464 (1892)) listed this as Pterygoloma repens (Hook.) 
Hanstein, which is figured in Botanical Magazine 72 : t. 4250 (1846) as Alloplectus 
repens from Colombia. Among Jamaican species of Columnea it is notable for its 
very slender creeping stems, small leaves and almost tubular narrow corolla with a 
short only slightly bilabiate limb. The type-specimen {Wullschlagel i2go, GOET) 
has red narrowly ovate sepals about 12 mm. long, 4-6 mm. broad at anthesis, and 
similar specimens have been collected in Westmoreland. Specimens collected by 
me in the Cockpit Country diverge in having green lanceolate sepals 8-5-10 mm. 
long, 3-5-4 mm. broad at anthesis. It grows on damp walls, moss-covered rocks 
and trees, in western Jamaica (Text-fig. 6). 

Hooker's Alloplectus repens, i.e. C. repens (Hook.) Hanstein (in Linnaea 34 : 395 
(1865)) was based on a cultivated plant said to have been introduced by Purdie 
from Colombia. It comes so close to C. jamaicensis, however, as to arouse suspicion 
that Hooker may have had an unusual variant of C. jamaicensis mislabelled as to 
origin. If no such plant as Hooker's occurs in Colombia, then the correct name for 
the Jamaican species will be C. repens. Some of Purdie's Colombian specimens have 
been attributed to Jamaica by Grisebach; hence the possibility of some of his 
Jamaican material being attributed to Colombia. 

2. Columnea hispida Swartz, Nov. Gen. & Sp. PI. : 94 (1788); Fl. Ind. Occ. 2 : 
1085 (1800). — Griseb., Fl. Brit. W. Ind. Is. : 465 (1862) excl. descript. — Urban, 
Symb. Antill. 2 : 364 (1901). — Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). 
(PI. 15; Text-figs. I A, 2 A, 7.) 

Shrub, the stems brownish grey, stout, hispid with stiff spreading hairs to 3 mm. 
long. Leaves paired, those of a pair markedly unequal, dark green above, paler green 
below, hirsute on the upper surface and at the margin with long erect reddish hairs 
to 4 mm. long, hirsute on the lower surface along the veins with similar long hairs 
and hispid on the surface itself with short colourless hairs to 0-5 mm. long, the larger 
leaf of the pair relatively long-stalked, asymmetric, the blade more or less elliptic or 
obovate, with the apex obtuse, the margin remotely and shallowly serrate or almost 
entire, the base asymmetric, its lower side broadly cuneate, its other side narrowly 



220 THE JAMAICANSPECIES OFCOLf/AfAT^:^ AND A LLOPLECT US 

cuneate and arising from the midrib about 5-10 mm. higher, 6-15 cm. long, 4-7 cm. 
broad, with four to six curved veins impressed each side of the midrib above, the 
smaller leaf 2-6 cm. long, I-5-3-5 cm. broad; petiole of larger leaf 1-2 cm. long, of 
smaller leaf 5-10 mm. long. Flowers up to three in a leaf-axil. Pedicels c. 5 mm. 
long, densely hispid. Calyx divided to the base, very pale green but hispid outside 
with long reddish hairs; segments very narrowly oblong, c. 15 mm. long in flower to 
2 cm. long in fruit, 3 mm. broad, entire. Corolla strongly 2-lipped, dull golden 
yellow, c. 2'5 cm. long, densely hispid on the back with hairs to 3 mm. long, the tube 
a little shorter than the calyx, c. i cm. long, the upper lip c. 1-5 cm. long and trilobed 
with a notched middle lobe and two rounded lateral lobes c. 4 mm. long, the lower 
lip c. I cm. long, 4 mm. broad, recurved. Stamens exserted; filaments minutely and 
sparsely hairy; anthers c. 2 mm. long. Ovary covered with long ascending bristles; 
style glandular-pubescent. 

St. James: White Rock Hill, one mile south of Sweetwater, 600 m., March 1956, 
Steam 44g (BM). Same locality, Feb. 1957, Proctor 161 58 (BM; IJ). Half mile 
south of Point P.O., 500-550 m.,1961, Proctor 22 5yy[\}). Near Catadupa {fide Morley). 

Unlocalized: Swartz (BM, type-collection). 

This species was described by Swartz from fruiting specimens collected by himself 
" in montibus altis Jamaicae occidentalis, ad rupes umbrosas " and it remained thus 
vaguely localized and not again collected, despite the wide-ranging activities of 
Purdie, Harris and Britton in Jamaica, until found by Proctor and myself in 1956, 
growing on shaded limestone cliffs in the west of the island (Text-fig. 7), as indicated 
by Swartz. The flowers were unknown until 1957. It has since been cultivated at 
Kew and Montreal botanic gardens. From other Jamaican species it is easily 
distinguished by the long reddish hairs on the upper surface of the leaves and by the 
small flowers. 

3. Columnea urbanii Stearn, sp. nov. (PI. 16; Text-figs. iB, 2H, 9, 28G.) 

Columnea hispida sensu Griseb., Fl. Brit. W. Ind. Is. : 465 (1862) pro parte maj.; non 

Swartz. 
Columnea hirsuta sensu Hanstein in Linnaea 34 : 413 (1865) pro parte quoad Wullschldgel 

gsg; non Swartz. ■ 

Columnea hirsuta var. subintegra f. wullschlaegelii Urban, Symb. Antill. 2 : 363 (1901). " 

Suffrutex, caulibus ascendentibus 4-6 mm. crassis, primo pilis patentibus ad 2 mm. 
longis hirsutis, demum griseis glabratis, internodiis 0-5-4 cm. longis, cicatricibus 
prominentibus. Folia ad apicem ramulorum conferta, per paria aequalia, subsessilia, 
plerumque late elliptica sed interdum rotunda vel late obovata, apice obtusa vel 
rotundata, margine subintegra, basi late cuneata raro asymmetrica, 3-8 cm. longa, 
2 -5-5 '5 cm, lata, in vivo utrinque pallide viridia, nervis lateralibus utroque latere 
costae 4-5 arcuatis, supra pilis erectis incoloribus plerumque 5-cellularibus c. 2 mm. 
longis hirsuta, subter pilis plerumque multo brevioribus appressis c. 0-5 mm. longis 
strigulosa; petiolus c. 1-4 mm. longus. Flores axillares 1-2. Pedicelli c. 8-15 mm. 
longi, hirsuti. Calyx viridis, extus hirsutus; segmenta libera subaequalia lanceolata 
acuta, anthesi 10-15 mm. longa, post anthesin ad 25 mm. longa, 3-5-5 mm. lata, 
integra. Corolla valde bilabiata, c. 5 cm. longa, flava, dorso pilis patentibus ad 



THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS 221 

2'5 mm. longis hirsuta; tubus basi postice saccatus gradatim ampliatus c. 22 mm. 
longus calyce fere duplo longior; labium superum c. 18-25 mm. longum, trilobatum, 
lobo medio emarginato c. 10 mm. lato, lobis lateralibus anguste triangularibus 
patentibus; labium inferum c. 18 mm. longum, 3 mm. latum, recurvatum. Stamina 
exserta, filamentis glabris, antheris oblongis c. 1-5-2 mm. longis. Ovarium apice 
hirsutum; stylus pubescens. 

Shrub, the stems ascending, grey, stout, 4-6 mm. thick, hirsute, with spreading 
hairs to 2 mm. long when young, later glabrous. Leaves paired, those of a pair 
almost equal, short-stalked or almost sessile, the blade mostly broadly elliptic but 
varying from rotund to broadly obovate, with the apex obtuse or rounded, the margin 
almost entire, the base broadly cuneate, sometimes asymmetric, 4-12 cm. long, 
2*5-5"5 cm. broad, light green on both surfaces, with 4-5 curved veins each side of 
the midrib impressed deeply above, hirsute above with erect mostly 5-celled hairs 
c. 2 mm. long, strigose below with numerous short hairs c. 0-5 mm. long, together 
with some longer hairs; petiole 1-4 mm. long. Flowers i or 2 in a leaf-axil. 
Pedicels c. 8-15 mm. long, hirsute. Calyx divided to the base, green, densely 
hirsute outside, less so inside; segments lanceolate, acute, 10-15 mm. long in flower, 
3-5 mm. broad, to 25 mm. long in fruit, entire. Corolla strongly 2-lipped, pale 
yellow, c. 5 cm. long, hirsute on the back with long spreading hairs to 2-5 mm. long, 
the tube nearly twice as long as the calyx, c. 22 mm. long, the upper lip c. 18-25 "irn- 
long, trilobed, with a notched middle lobe c. 10 mm. broad and two spreading 
narrowly triangular acute lateral lobes, the lower lip c. 18 mm. long, 3 mm. broad, 
recurved. Stamens exserted; filaments glabrous; anthers c. 1-5-2 mm. long. Ovary 
hirsute at the apex; style pubescent. 

Manchester: Top Hill, 820 m., March 1966, Morley 325 (BM). Fairfield, 1849, 
Wullschldgel g^g (GOET; M; type-collection of C. hirsuta var. subintegra f. wull- 
schlaegelii) . Along road between Fairfield and Huntley, 800-900 m., Jan. 1957, 
Proctor i6iog (IJ). Huntley, 900 m., 25 Feb. 1956, Steam 361 (BM, holotype). 
Proctor 11621 (IJ). Near Christiana, May 1901, Fawcett 8145 (UCWI). Upper 
slopes and summit of Heron's Hill (Shooter's Hill, Martin's Hill), 900-940 m., 
March 1958, Prodor I748g (BM; IJ). Banana Ground, 900 m., Dec. 1961, Adams 
ioi2y (UCWI). Oatlands, north-west of Rose Hill, 830 m., Jan. 1961, Proctor 
2igi5 (BM). 

St. Ann: Mason River district, three to four miles northwest of Kellits, 760m., 
March 1957, Proctor 16201 (BM; IJ). 

Clarendon: Peckham Woods, 760m., March 1954, Proctor 8422 (I J). Knox 
College, Spaldings, 850 m., March 1952, Proctor 6345 (IJ). 

This species was first collected by the Moravian missionary Wullschlagel in 1849 
and identified by Grisebach with the very different C. hispida. Urban recognized 
its distinctness but, having only Wullschlagel's limited material at hand, described it 
briefly as a form of C. hirsuta, which is confined to the Blue Mountains of Jamaica. 
In a living state the leaves are a bright green with the veins deeply impressed above, 
and the flowers are pale yellow. It grows among both rocks and trees, and sometimes 
on stone walls, in central Jamaica (Text-fig. 9). Fairfield, where Wullschlagel 
collected his specimens, and Huntley, where I collected mine, are about five miles 



222 THE JAMAICANSPECIESOFCOLf/MATf^ A^D A LLOPLECT US 

apart and near the western boundary of Manchester. It is named in honour of Ignaz 
Urban (1848-1931), the most scholarly, industrious and scientifically prolific of the 
many botanists who have studied West Indian plants; a short survey of his career 
by R. A. Howard will be found in E. Carroll and S. Sutton, Cumulative Index to the 
nine Volumes of the Symbolae Antillanae : 1-5 (1965). Among hybrids between this 
species and C. rutilans are some plants clearly intermediate, others approximating 
more closely to one or other parent species, probably as a result of back-crossing. 
Those nearest to C. urhanii in leaf -size and pubescence may nevertheless usually be 
distinguished from it by having the base of the leaf-blade more attenuate and the 
corolla partly red, although sometimes they agree with it in almost all characters 
but have only minute bristles on the upper surface of the leaf. 

4. Columnea proctorii Stearn, sp. nov. (PI. 17, 18; Text-figs. 2B, 8.) 

Suffrutex, caulibus ascendentibus c. 5 mm. crassis, pilis patentibus ad 3 mm. 
longis hispidis, internodiis longis, cicatricibus prominentibus. Folia ad apicem 
5-14 cm. longum ramulorum disposita, per paria valde inaequalia, folio majore 
breviter petiolato asymmetrico, lamina ovata ad obovata, apice acuta vel breviter 
acuminata, margine subintegra, basi uno latere rotundato, altero cuneato, 5-10 cm. 
longa, 2'5-6-5 cm. lata, pallide viridia, nervis lateralibus utroque latere costae 4-6 
arcuatis supra profunde impressis, utrinque pilis erectis incoloribus plerumque 
5-cellularibus I-5-2-5 mm. longis hirsuta, folio minore fere sessili orbiculato 1-3-5 cm. 
longo latoque, similiter vestito; petiolus folii majoris 2-8 mm. longus. Flores 
axillares 1-3. Pedicelli 12-20 mm. longi, dense hirsuti. Calyx viridis, extus pilis 
longis glandulosis hirsutus; segmenta libera subaequalia anguste lanceolata acuta, 
anthesi c. 2 cm. longa, post anthesin c. 3 cm. longa, 2-3 mm. lata, Integra. Corolla 
valde bilabiata, c. 4-5 cm. longa, flava rubrovittata, dorso pilis patentibus glandulosis 
ad 2 mm. longis dense hirsuta; tubus basi postice saccatus gradatim ampliatus 
c. 18-25 mm. longus calycem aequans; labium superum c. 18 mm. longum, trilo- 
batum, lobo medio emarginato c. 7 mm. lato, lobis lateralibus anguste triangularibus 
vel fere lanceolatis obtusis patentibus; labium inferum c. 12 mm. longum, 4 mm. 
latum, recurvatum. Stamina exserta, filamentis glabris, antheris oblongis 2 mm. 
longis. Ovarium basi glabrum, supra medium pubescens; stylus glanduloso- 
pubescens. 

Shrub, the stems grey, stout, c. 5 mm. thick, hispid with spreading hairs to 3 mm. 
long. Leaves occupying the terminal 5-14 cm. of the shoots, paired, those of a pair 
markedly unequal, light green, hirsute on both surfaces with erect colourless hairs 
I-5-2-5 mm. long, the larger leaf of the pair short-stalked, asymmetric, the blade 
more or less ovate to obovate, with the apex acute or shortly acuminate, the margin 
almost entire, the base asymmetric, its lowest side rounded, the other cuneate and 
arising from the midrib about 5 mm. higher, 5-10 cm. long, 2-5-6-5 cm. broad, with 
four to six curved veins each side of the midrib impressed deeply above, the smaller 
leaf almost sessile and orbicular, 1-3-5 cm. long and broad; petiole of larger leaf 
2-8 mm. long. Flowers 1-3 in a leaf-axil. Pedicels 12-20 mm. long, hirsute. Calyx 
divided to the base, light green, hirsute with long glandular hairs, the segments 



THE J AM A IC AN S PEC I ES OF COL C7MA^£ J A-ND A LLOPLECT US 223 

narrowly lanceolate, acute, c. 2 cm. long, 2-3 mm. broad in flower to 3 cm. long in 
fruit, entire. Corolla strongly 2-lipped, pale yellow with longitudinal red stripes, 
c. 4-5 cm. long, densely hirsute on the back with long spreading glandular hairs to 
2 mm. long, the tube almost as long as the calyx, c. 18-25 ^nni- long, the upper lip 
c. 18 mm. long, trilobed, with a notched middle lobe about 7 mm. broad and two 
spreading narrowly triangular to almost lanceolate obtuse lateral lobes, the lower lip 
c. 12 mm. long, 4 mm. broad, recurved. Stamens exserted; filaments glabrous; 
anthers 2 mm. long. Ovary glabrous in the lower part, pubescent in the upper part 
style glandular-pubescent. 

Hanover: Ramble, March 1905, Fawcett 8gi6 (UCWI). 

St. James: Chatsworth district one and a half miles north-north-west of Maroon 
Town, 480 m., March 1956, Steam 451 (BM, holotype; K), Proctor 11733 (IJ). White 
Rock Hill, one mile south of Sweet Water, 600-670 m., Feb. 1957, Proctor 16160 (IJ). 
Same locality, Dec. 1962, Proctor 2joog (BM; IJ). 

St. Elizabeth: Maggotty Falls, by the Black River, 120m., May 1956, Steam 
lojg (BM), Proctor 15382 (IJ). 

Trelawny: Cockpit Country at Black Banana Hole, about five miles north of 
Accompong, 330 m., April 1956, Steam g38 (BM). Cockpit Country, near Red Mud 
Hole, about six miles north-north-west of Accompong, 330 m., April 1956, Steam g^y 
(BM; K). Cockpit Country, at Ramgoat Cave, 460 m., Jan. 1956, Howard, Proctor 
& Steam 14645 (BM; IJ). Miss Laura's Hill, Wilson Valley district, one mile north 
of Warsop, 600-670 m., Jan. 1964, Proctor 24452 (BM; IJ). Island View Hill, 
Wilson Valley district, one and a half miles north of Warsop, 600-670 m., March 
i960. Proctor 20756 (BM; IJ). Five miles northwest of Troy, 530 m., Sept. 1963, 
Adams 12420 (BM). 

Manchester: Half a mile northwest of Christiana, 900 m., Oct. 1958, Proctor 
18288 (BM; IJ). Same locality, Feb. i960. Proctor 205g5 (BM; IJ). One mile 
southeast of Pike, 900 m., Oct. 1958, Proctor 18308 (BM; IJ). 

This species is widespread in the Cockpit Country and adjacent areas of western 
Jamaica (Text-fig. 8) and is apparently more northern in distribution than C. urbanii, 
from which it differs in having the leaves of a pair markedly unequal in size and 
hirsute below, and the corolla usually striped longitudinally with red; the five free 
calyx-segments distinguish C. proctorii from C. suhcordata. I have named it in honour 
of George Richardson Proctor, author of A Preliminary Checklist of Jamaican 
Pteridophyta (1953), Additions to the Flora of Jamaica (1967) and other publications 
on Jamaican plants, whose collecting activities, extended since 1949 over almost the 
whole of Jamaica, have so notably increased knowledge of the distribution of species 
within the island and in whose company in 1956 I had the privilege of collecting the 
type material of this species. 

5. Columnea subcordata Morton in Contr. U.S. Nat. Herb. 29 : 6 (1944). — 
Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). (PI. 19; Text-figs. 2G, 10.) 

Shrub, the stems grey or brownish, stout, c. 5 mm. thick, hispid when young with 
long mostly spreading brownish multicellular hairs c. 2-3 mm. long. Leaves paired, 
those of a pair unequal, all short-stalked, green and hirsute with sparse long whitish 



224 THE JAMAICAN SPECIES OF COLUMNEA A'NDALLOPLECTUS 

multicellular hairs 0-8-I-8 m . long on both surfaces, the blade of the larger leaf 
of the pair more or less oblong-elliptic, with the apex abruptly and shortly acuminate, 
the margin entire and long-ciliate, the base subcordate and slightly oblique, 8-14 cm. 
long, 4'5-6-5 cm. broad, the blade of the smaller leaf (which usually falls quickly) 
broadly ovate or broadly obovate, 2-5-3-5 cm. long, 2 cm. broad; petiole 3-10 mm. 
long. Flowers 1-4 in a leaf-axil. Pedicels 1-5-2 cm. long, hirsute. Calyx 2-lipped, 
the four anterior segments joined from the base for 5 mm. (in flower) to 8 mm. but 
the posterior segment free to the base, light green, hispid with long whitish hairs on 
both surfaces, I-5-2-2 cm. long, the four free segments narrowly oblong-lanceolate 
2"5-4*5 mm. broad, acute, the posterior segment linear-tapering, all entire. Corolla 
strongly 2-lipped, light yellow, glandular-pilose, 4-4-8 cm. long, the tube mostly 
covered by the calyx, the upper lip c. 2-5 cm. long with a deeply notched middle lobe 
and two spreading oblong lateral lobes, the lower lip c. 1-5 cm. long, narrowly oblong, 
refiexed. Stamens exserted; filaments pubescent; anthers 2 mm. long. Ovary 
sericeous; style glandular-pilose. 

Trelawny: Mount Ridgway Road, Windsor, 100-150 m., April 1931, G. S. Miller 
1556 (US, type). Windsor House estate, 120 m., March 1956, Steam 4J2 (BM; K). 

Columnea subcordata stands apart from all other Jamaican species of Columnea in 
the fusion of the four anterior calyx-segments into one for the lower third of their 
length, with the posterior one remaining completely free. It probably arose as a 
mutation from C. proctorii. At the type-locality, Proctor and I found it growing 
high on trees, the sparingly branched stems hanging to a length of six feet. The 
leaves are dark green above, the hairs colourless, the corolla light yellow (Chinese 
yellow). It is evidently confined to a small area of western Jamaica (Text-fig. 10). 

6. Columnea harrisii (Urban) Britton ex Morton in Contr. U.S. Nat. Herb. 29 : 10 
(1944). — Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). (Text-fig. lE, 11.) 
Columnea hirsuta var. subiniegra f. harrisii Urban, Symb. Antill. 2 : 363 (1901). 

Shrub, the stems grey, fairly slender, pilose with whitish hairs. Leaves paired, 
those of a pair unequal or almost equal, green and sericeous with numerous slightly 
arched more or less appressed multicellular hairs on both surfaces, the larger leaf 
of an unequal pair asymmetric, the blade more or less narrowly elliptic-oblong, with 
the apex shortly acuminate, the margin remotely and shallowly toothed, the base 
rounded on the outer side, acute on the inner side, 7-5-13 cm. long, 3-5 cm. broad, 
the smaller leaf broadly or narrowly ovate or almost orbicular, 1-5-4-5 cm. long, 
1-0-2-3 cm. broad; petiole of larger leaves 3-8 mm. long, of smaller leaves 1-3 mm. 
long. Flowers i to 3 in a leaf axil. Pedicels 5-25 mm. long, hirsute. Calyx divided 
to the base, green, densely sericeous on both surfaces ; segments lanceolate to narrowly 
lanceolate, 1-8-2-5 cm. long, c. 4 mm. broad, entire or with one or two minute teeth 
on each margin. Corolla strongly 2-lipped, yellow striped with red, 3-8-5 cm. long, 
sericeous outside with long whitish hairs, the tube almost covered by the calyx, the 
upper lip c. 2-3 cm. long with a notched middle lobe and two spreading narrowly- 
oblong lateral lobes, the lower lip c. 12-15 mm. long, narrowly oblong, recurved. 
Stamens exserted; filaments glabrous; anthers c. i-8 mm. long. Ovary sericeous; 
style glandular-pilose. 



► 



THE JAMAICAN SPECIES OF COL f/MiV£^ AND ^ LLOPLECT C/S 225 

Hanover: Belvedere, 160 m., Jan. 1899, Harris ^^22 (BM, isotype). Same 
locality, March 1903, Fawcett 8480 (BM; NY). Hills near Kempshot, March 1908, 
Britton 2438 (NY; US). Shettlewood, 350-450 m., Feb. 1961, Proctor & Mullings 
22016 (BM; IJ). 

Westmoreland: Newmarket, Sept. 1907, Britton I5y6 (NY). Woodstock, 
400 m., Aug. 1907, Harris (UCWI). Same locality, 360-420 m., Oct, i960, Proctor 
2 1 519 (BM; IJ). 

St. James: One mile southwest of Montpelier, no m., Feb. 1961, Proctor & 
Mullings 22010 (BM; IJ). 

Columnea harrisii, although placed by Urban under C. hirsuta, is abundantly 
distinct in its sericeous pubescence, entire calyx-segments and glabrous stamens, its 
distribution, and its ecology, as pointed out by Morton. It is a lowland species 
confined to the west of the island (Text-fig. 11), where it is epiphytic on large trees 
and often associated with bromeliads. Its name commemorates William Harris 
(1860-1920), for many years Superintendent of Public Gardens in Jamaica, whose 
extensive collections from 1890 onwards led to the discovery in Jamaica of 
approaching 400 species above those recorded for the island in Grisebach's Flora 
(1859-64). 

7. Columnea argentea Griseb., Fl. Brit. W. Ind. Is. : 465 (1862). — Urban, 
Symb. Antill. 2 : 361 (1901). — Morton in Contr. U.S. Nat. Herb. 29 : 9 (1944). — 
Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). (PI. 20; Text-figs. iF, 2C, 
5, 12.) 
C. wullschlaegeliana Hanstein in Linnaea 34 : 414 (1865). 

Shrub, the stems stout, c. 5 mm. thick, silky at the apex with white hairs. Leaves 
paired, those of a pair almost equal, the blade very narrowly elliptic, narrowly 
elliptic or oblanceolate, with the apex shortly acuminate, the margin remotely and 
shallowly toothed, almost entire, the base rounded or cuneate, 7-5-17 cm. long, 
2-5-4 cm. broad, grey-sericeous on both surfaces; petiole 4-10 mm. long. Flowers 
I to 4 in a leaf axil. Pedicels 5-25 mm. long, sericeous. Calyx divided to the base, 
densely sericeous on both surfaces; segments equal, oblong-linear, c. 14-18 mm. in 
flower to 20-24 Kiii^- long in fruit, 2-4 mm. broad, entire. Corolla strongly 2-lipped, 
pale yellow, 4-5 cm. long, densely white-sericeous outside, the tube slightly longer 
than the calyx, c. 2-0-2-5 cni. long, the upper lip c. 2-0-2-5 cm long, with a notched 
middle lobe and two narrowly triangular lateral lobes, the lower lip 15-18 mm. long, 
oblong-linear, reflexed. Stamens exserted; filaments glabrous; anthers 1-5 mm, 
long. Ovary sericeous; style glandular-pilose. Berry ellipsoid, pink, c. 11 mm. 
long, 8 mm. broad. 

Manchester: Nazareth, 1849, Wullschldgel g^g (GOET, holotype of C. argentea; 
M, holotype of C. wullschlaegeliana). Nazareth, among rocks behind the Moravian 
Church, one mile south of Maidstone P.O., 880 m., Feb. 1956, Stearn J59 (BM; K; 
topotype of C. argentea and C. wullschlaegeliana), Proctor 11615 (IJ). Alumina 
Jamaica property near Kirkvine Works, Shooters Hill, 380-460 m., Jan. 1958, 
Howard & Proctor I4g44 (BM ; IJ). Bath district, along Somerset Road southeast of 



226 THE JAM AI CAN SPECIES OF COL f7MN£ J Ai^DALLOPLECTUS 

Johns Hall, 700 m., Jan. 1961, Proctor 21883 (BM; IJ). New Green, Sept. 1908, 
Britton 3755 (NY). Clones, Ballynure to Malton, 700 m., April 1962, Adams 11018 
(BM). 

Columnea argentea is an attractive and easily recognized species, notable for the 
silvery and silky pubescence which clothes the relatively long and narrow leaves, the 
calyx and the outside of the pale clear yellow corolla. At the type-locality I found it 
growing among rocks. Howard & Proctor record it as epiphytic on clumps of 
Hohenbergia in trees. All collections have come from within ten miles of one another 
in central Manchester (Text-fig. 12). 

8. Columnea brevipila Urban, Symb. Antill. 6 : 41 (1909).^ — ^Morton in Contr. 
U.S. Nat. Herb. 29 : 9 (1944). — ^Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). 
(Text-fig. iD, 13.) 

Shrub, the stems grey, stout, 5-8 mm. thick, strigillose when young with minute 
white appressed hairs and with lines of more villous hairs between the leaf-insertions, 
later glabrous. Leaves paired, those of a pair almost equal, shortly stalked, the blade 
narrowly oblong-elliptic or narrowly elliptic, with the apex shortly acuminate, the 
margin remotely and obscurely toothed, the base cuneate, 5-1 1 cm. long, 2-5 cm. 
broad, green and strigillose on both surfaces with minute whitish appressed hairs 
c. 0-2 mm. long; petiole 7-1 1 mm. long. Flowers i to 5 in a leaf-axil. Pedicels 
1-2 cm. long, strigillose. Calyx divided to the base, green, minutely strigillose on 
both surfaces; segments narrowly lanceolate, acute, 2-3 cm. long, 3-5 mm. broad, 
entire. Corolla strongly 2-lipped, completely yellow, 4-5-5 cm. long, sericeous out- 
side with long whitish hairs, the tube mostly covered by the calyx, the upper lip c. 
2-2-5 cm. long, with a strongly notched middle lobe and two spreading and ascending 
oblong lateral lobes, the lower lip to 1-5 cm. long, narrowly oblong, reflexed. Stamens 
exserted; filaments glabrous; anthers 2 mm. long. Ovary densely hirsute; style 
glandular-pilose . 

Westmoreland: Kentucky Hill, Bluefields Mountains, 600-760 m., March 1908, 
Harris loigg (NY; UCWI; US; type-collection). Hills west of Bognie, Bluefields 
Mountains, 700-760 m., Feb. 1961, Proctor & Mullings 22043 (BM). Teague Gully, 
300-400 m., Feb. 1961, Proctor & Mullings 22062 (BM). Cairn Curran district, 
two and a half miles west-north-west of Hopewell, 530 m., March jg^6, Stearn 433 
(BM; K), Proctor 11719 (IJ). 

St. Elizabeth: One mile north of Redgate, 150 m., March 1956, Steam 438 (BM). 

Columnea brevipila agrees with C. fawcettii in having the upper surface of the leaf 
beset with minute appressed hairs but differs completely in its entire sepals. It is 
confined to southwestern Jamaica (Text-fig. 13). Some hybrids between C. rutilans 
and C. urbanii come close to it in general character but differ in having their pedicels 
and sepals clothed with spreading hairs. 

9. Columnea hirsuta Swartz, Nov. Gen. & Sp. PI. : 94 (1788); Fl. Ind. Occ. 2 : 
1080 (1800) excl. syn. — Hook, in Curtis, Bot. Mag. 58 : t. 3081 (1831).— Griseb., 
Fl. Brit. W. Ind. Is. : 465 (1862) excl. parte. — Hanstein in Linnaea 34 : 413 



THE JAMAICAN SPECIES OF COLUMNEA AND A L LOP LECT US 227 

(1865) excl. parte. — Urban, Symb. Antill. 2 : 361 (1901) excl. var. subintegra 
& var. fawcettii. — Morton in Contr. U.S. Nat. Herb. 29 : 7 (1944). — Stearn in 
Proc. Linn. Soc. Lond. 170 : 141 (1959). (Text-fig. iC, 14.) 

Eusynetra bicolor Raf., Fl. Tellur. 2 : 57 (1837), nom. superfl. 
Columnea hirsuta var. genuina Urban, torn. cit. : 361 (1901). 
Columnea hirsuta var. pallescens Urban, torn. cit. : 362 (1901). 
Columnea hirsuta var. concolor Urban, torn. cit. : 362 (1901). 

Shrub, the stems brownish grey, moderately slender, densely hirsute when young 
with long multicellular hairs. Leaves paired, those of a pair unequal, green and 
hirsute on both surfaces with long multicellular flaccid hairs together with short 
2-3-celled hairs, the blade of the larger leaf oblong-elliptic to narrowly obovate, with 
the apex acute or shortly acuminate, the margin shallowly serrate, the base rounded 
or cuneate, 4-5-12 cm. long, 2-5-5-5 cm. broad, the blade of the smaller leaf elliptic 
or broadly elliptic, 2-6 cm. long, I-8-3-8 cm. broad; petiole c. 15 mm. long. Flowers 
1-3 in a leaf-axil. Pedicels 1-2 cm. long, hirsute. Calyx divided to the base, red or 
green, densely hirsute on both surfaces with long multicellular red or colourless hairs ; 
segments linear, slightly uneven, c. 2 cm. long in flower to 4 cm. long in fruit, 2-4 cm. 
broad, remotely toothed with one to three teeth on each margin. Corolla strongly 
2-lipped, red with broad yellow longitudinal stripes, c. 4-5-5 cm. long, the tube mostly 
covered by the calyx, c. 2 cm. long, the upper lip c. 2-5 cm. long, with a deeply 
notched middle lobe and two spreading narrowly triangular lateral lobes, the lower 
lip c. 2-2 cm. long, oblong-linear, reflexed. Stamens exserted; filaments sparsely 
pubescent; anthers i •8-2-5 mm. long. Ovary sericeous; style densely glandular- 
pilose. Berry white. 

St. Andrew: Moresham River, quarter mile above Hermitage Dam, 350 m., 
March 1956, Stearn 412 (BM). Hardwar Gap, 1,200 m., Sept, 1952, Robbins 1511 
(UCWI). Summit of Catherine's Peak, 1,540 m., Nov. 1952, Proctor ^341 (IJ). 
West slope of Catherine's Peak, 1,300 m., Jan. 1961, Proctor 2ig$4 (BM). 
Catherine's Peak, Jan. 1888, Eggers 3603 (US, type-collection of C. hirsuta var. 
pallescens). Vicinity of St. Helen's Gap, 1,475 m., March 1920, Maxon & Killip ^gi 
(US). Morces Gap, 1,520 m., July 1903, Nichols 15 (K; US). Same locality, 
1,500 m., March 1920, Maxon & Killip i'j22 (US). Along trail between Cinchona 
and Morces Gap, 1,520 m., Feb. 1941, Burrowes 3563 (IJ). Near Cinchona, 1,600 m.. 
June 1896, Fawcett 6366 (BM, type-collection of C. hirsuta var. concolor). Cinchona, 
windward slopes, Feb. 1915, Harris & Lawrence C 151 58 (US). 

Portland: Trail from Morces Gap to Vinegar Hill, 1,175-1,500 m., March 1920, 
Maxon & Killip 6g4 (US). Near Mabess River, 780 m., Dec. 1905, Harris gi23 (NY). 
Lower eastern ridge of Mossman's Peak, 1,600-1,700 m.. May 1926, Maxon g665 
(US). East slope of the John Crow Mountains two miles southwest of Ecclesdown, 
600 m., March 1951, Proctor 5yo3 (IJ). Same locality, Jan. 1956, Howard, Proctor 
& Stearn i4ygo (BM). Upper north slope of Silver Hill Gap, along track toward 
Big Level, 1,100 m., Feb. 1963, Proctor 23255 (BM; IJ). 

St. Thomas: Near Whitfield Hall, 1,400 m., Jan. 1956, Stearn 123 (BM). Portland 
Gap, 1,640 m., Dec. 1954, Proctor g6i5 (BM). Same locality, Jan. 1956, Stearn g4 
(BM). Northwest slope of Yallahs Hill, 600 m., April 1956, Stearn 6yy (BM), 

BOX. 4, 5. 14 



228 THE JAMAICAN SPECIES OF COLUMNEA ANBALLOPLECTUS 

Unlocalized : 1777, Shakespear (BM). W. Wright (BM). Swartz (BM, type- 
collection of C. hirsuta). 1843, Lane (K). Purdie (K). March (K). 

An eastern Jamaican species characteristic of the Blue Mountains (Text-fig. 14), 
Columnea hirsuta is easily distinguished from its closest ally, C. fawcettii, by the long 
multicellular hairs on the upper leaf-surface. The leaves green beneath separate 
these species at a glance from C. rutilans, which has also more deeply toothed calyx- 
segments. 

Usually C. hirsuta grows as an epiphyte, the vegetative shoots clinging like ivy to 
mossy tree trunks while the flowering shoots spread horizontally outwards, but it 
may sometimes be found growing terrestially among rocks. 

10. Columnea fawcettii (Urban) Morton in Contr. U.S. Nat. Herb. 29 : 8 (1944). — 
Stearn in Proc. Linn. Soc. Lond. 170 : 141 (1959). — Raymond in Mem. Jard. 
Bot. Montreal 57 : 13, t. 6 (1965). (Text-fig. iH, 2D, 15, 28.) 

Columnea hirsuta sensu Griseb., Fl. Brit. W. Ind. Is. : 465 (1862) pro parte; non Swartz, 
Columnea hirsuta vai. fawcettii Urban, Symb. Antill. 2 : 363 (1901). 

Rapunculus fruticosus foliis oblongis integris villosis etc. Sloane, Cat. PL Jam. : 58 (1696); 
Voy. Nat. Hist. Jam. i : 157, t. 100 f. i (1707). 

Shrub, the stems grey, fairly slender, pilose with whitish hairs. Leaves paired, 
those of a pair equal to very unequal, green and shortly strigillose on both surfaces, 
the hairs of the upper surface 2-celled, 0-3-0-4 mm. long, the blade of the larger leaf 
mostly elliptic or narrowly elliptic, with the apex shortly acuminate, the margin 
remotely and obscurely toothed, the base asymmetric, cuneate, 6-10-5 cm. long, 
2-5-4-5 cm. broad, the blade of the smaller leaf in shape like the larger, 3-4-5 cm. 
long; petiole of larger leaves 8-16 mm. long, of smaller 2-4 mm. long. Flowers 
I to 3 in a leaf-axil. Pedicels 12-15 mm. long, hirsute. Calyx divided to the base, 
red, densely hirsute on both surfaces; segments slightly uneven, narrowly lanceolate, 
2-3-2-8 cm. long, 3-4 mm. broad, remotely toothed, with two or three linear teeth 
to 2 mm. long on each margin. Corolla strongly 2-lipped, red with yellow longi- 
tudinal stripes, c. 4-5-5 cm. long, sericeous with long whitish hairs, the tube covered 
by the calyx, the upper lip c. 2-8 cm. long, with a strongly notched middle lobe and 
two spreading narrowly triangular lateral lobes, the lower lip c. 2 cm. long, narrowly 
oblong, reflexed. Stamens exserted; filaments slightly pubescent; anthers c. 1-5 mm. 
long. Ovary sericeous; style glandular-pilose. Berry ovoid, white. 

St. Ann : Interior Road, near Browns Town, Feb. 1897, Fawcett 6587 (BM, type- 
collection). Dry Harbour Mountains, two miles west of Albion, 760 m., Jan. 1956, 
Stearn i^y (BM). Vicinity of Lydford Post Office, Governor's Hill, 560 m., Dec. 
1953, Howard & Proctor 13570 (IJ). Reynolds Mine Area near Lydford Post Office, 
360-420 m., Jan. 1956, Howard, Prottor & Stearn 14602 (BM). Averham Park, 
560 m., March 1956, Stearn 585 (BM). Soho, 430 m., May 1915, Harris 11993 (NY). 
St. Ann's Road, four miles from Moneague, Dec. 1849, Alexander (K). Union Hill, 
near Moneague, April 1908, Britton & Hollick 2741 (NY). Grierfield, near Moneague, 
April 1908, Britton 2659 (NY). Lower part of road to Holly Mount, 400 m., March 
1956, Stearn 5y6 (BM). Near Camperdown School, one and a half miles southwest 
of Gibralter, 420 m., March 1956, Stearn 568 (BM; K). 



THE JAMAICAN SPECIES OF COLUMNEA K^T> ALLOP LECT U S 229 

St. Catherine: Vicinity of Holly Mount, Mount Diablo, 750m., May 1904. 
Maxon 2334 (US). Same locality, Feb. 1905, Harris 88g6 (K; NY). Mount 
Diablo, Feb. 1916, Ridley (K). Upper slopes of Mount Diablo, 500-800 m., Feb. 
1920, Maxon & Killip 42g (NY; US). Between Point Hill and Juan de Bolas Peak, 
610-820 m., March 1958, Yuncker i83y4 (BM). 

St. Andrew: Cooper's Hill, Red Hills, 670-760 m., March 1954, Proctor 8488 (IJ; 
NY). Same locality, Feb. 1956, Steam 25s (BM; K). 

Portland: Port Antonio-Moore Town Road, March 1924, Norman 225 (BM). 
Uncommon Hill, above Fruitful Vale Post Office, 670-760 m., April 1954, Proctor 
8579 (I J)- Seamen's Valley, 150-250 m., Feb. 1920, Maxon & Killip 81 (NY; US). 
Alligator Church district, one mile south of Seamen's Valley, 70-150 m., March i960. 
Proctor 20640 (BM). Northwest slope of Joe Hill, 300-670 m., April 1955, Proctor 
10102 (IJ). Same locality, Feb. 1956, Steam 243 (BM). Vicinity of Millbank, 
200-300 m., Feb. 1920, Maxon & Killip 133 (NY; US). East slope of John Crow 
Mountains, one to one and a half miles southwest of Ecclesdown, 300-460 m., April 
1954, Proctor 8603 (IJ). Same locality, 460-600 m., Aug. 1954, Webster ^^yg 
(BM). Same locality, Jan. 1955, Proctor gyg8 (IJ). Ten miles southwest of 
Priestmans River, 460-760 m., Jan. 1951, Proctor 5255 (IJ). 

St. Thomas: Maccasucker Bump, 825-1,025 m., June 1926, Maxon g523 (US). 
Ridge east of Cuna Cuna Gap, 750-840 m., June 1926, Maxon g4oy (US). North 
slope of Corn Puss Gap, 600 m., Jan. 1945, Barry 1038 (I J). Sulphur Creek, above 
Fountain Inn [Bath Fountain], 150-200 m., Nov. 1957, Yuncker 1^486 (BM). 

Unlocalized: Sloane, Herb. 3 : fol. 21 (BM). 1859-60, Metcalfe (BM). Purdie (K). 

Columnea fawcettii is the most widespread Jamaican species of Columnea (Text-fig. 
15) and was the first to be collected and recorded, having been gathered by Sir Hans 
Sloane as long ago as 1688-89 ^^^ illustrated by him, albeit crudely, in 1707. Never- 
theless, because it closely resembles C. hirsuta, it remained without varietal or 
specific designation until 190 1, when Urban named it C. hirsuta var. fawcettii; not 
indeed until 1944 was its true status recognized, Morton then giving Urban's variety 
specific rank as C. fawcettii. It differs from C. hirsuta principally in having only 
short appressed hairs on the upper surface of the leaves. Its name commemorates 
William Fawcett (1851-1926), from 1880-86 assistant botanist at the British Museum, 
a period during which the botanical collections were transferred from Bloomsbury 
to South Kensington, then from 1887 until his retirement in 1908 Director of Botanic 
Gardens and Plantations in Jamaica. When in Jamaica he became especially 
interested in the Orchidaceae of the island and his monograph, prepared in collabora- 
tion with A. B. Rendle (1865-1938), was converted into the first volume (published 
in 1910) of the Flora of Jamaica. He worked continuously on this Flora at the 
British Museum (Natural History) until his death in 1926. 

II. Columnea rutilans Swartz, Nov. Gen. & Sp. PI. : 94 (1788); Fl. Ind. Occ. 
2 : 1083 (1800).— DC, Prodr. 7 : 542 (1839).— Griseb., Fl. Brit. W. Ind. Is. : 465 
(1862). — Hanstein in Linnaea 34 : 415 (1865). — Urban, Symb. Antill. 2 : 363 
(1901). — Morton in Contr. U.S. Nat. Herb. 29 : 5 (1944). — Stearn in Proc. Linn. 
Soc. Lond. 170 : 141 (1959). (PI. 21; Text-figs. iG, 2F, 17.) 
BOT. 4, 5. i4§ 



230 THE JAMAICANSPECIES OFCOLC/MAT^yl A-ND A LLOPLECT US 

Shrub, the stems grey, stout, when young strigose or almost tomentose with 
reddish hairs, later sparingly strigose. Leaves paired, those of a pair unequal, the 
upper surface dark green and strigillose with short multicellular hairs, the lower 
surface red-veined or completely red, pilose and almost sericeous, the larger leaf of 
the pair long-stalked, the blade lanceolate or narrowly lanceolate, with the apex 
shortly acuminate, the margin remotely serrulate, the base asymmetric, acute, 
7-18 cm. long, 2-5-6 cm. broad, smaller leaf of the pair soon falling, short-stalked, 
the blade 2-4-5 cm. long, 1-5-2-5 cm. broad; petiole of larger leaves 1-2-5 cm. long, of 
smaller leaves 3-5 mm. long, sericeous. Flowers i to 4 in a leaf-axil. Pedicels 
1-2 cm. long, strigillose. Calyx divided to base, pilose on both surfaces with long 
red hairs; segments 2-2-5 cm. long in flower to 3-5 cm. long in fruit, pinnately cleft 
with three acute teeth on each margin. Corolla strongly 2-lipped, red with longi- 
tudinal orange stripes, 4-5-5 -5 cm. long, the tube covered by the calyx, the upper 
lip to 3 cm. long, with a strongly notched galea and two spreading narrowly triangular 
lateral lobes, the lower lip to 2-5 cm. long, narrowly oblong, reflexed. Stamens 
exserted; filaments glabrous or slightly hairy, pale yellow; anthers 2-5 mm. long. 
Ovary sericeous; style glandular-pilose. Berry white, globose, c. 12 mm. broad, 
pilose. 

Hanover: Northwestern slope of Dolphin Head, May 1906, Harris g265 (NY; 
US). Interior summit slopes of Dolphin Head, 460-520 m., April 1955, Proctor 
10036 (BM; IJ). Same locality, March i960. Proctor 20689 (BM; IJ). 

St. James : Chatsworth district, one and a half miles north-north-west of Maroon 
Town, 490 m., March 1956, Steam 433 (BM). 

St. Elizabeth: Near Accompong, 400 m., April 1956, Steam 1003 (BM). Cooks 
Bottom, east of Elderslie, 490-520 m.. May i960. Proctor 20864 (BM). 

Trelawny: Near Troy, 600 m., April 1906, Harris g3yo (BM; K; NY). Same 
locality, 600-660 m., June 1906, Maxon 2852 (NY; US). Bartley's Run, three 
miles north-north-west of Troy, 520 m., April 1956, Steam g2y (BM). Tyre district, 
two miles north of Troy, 530 m., March 1955, Proctor gg26 (IJ). Six miles from 
Troy, April 1954, Robertson 1058 (UCWI). 

Manchester: Feb, 1844, Purdie (K). Near Christiana, May 1901, Fawcett 8144 
(UCWI). Heron's Hill (Shooter's Hill, Martin's Hill), 910-940 m., March 1958, 
Proctor 17482 (BM). Top Hill, 820 m., March 1966, Morley 344 (BM). 

Clarendon: Peckham Woods, 750 m.. May 1912, Harris iiog2 (NY; US). Same 
locality, March 1954, Proctor 8424 (IJ). Same locality. May 1955, Robertson 2034 
(K). Knox College, Spaldings, 850 m., May 1949, Dignum 34 (IJ). Bull Head, 
March 1903, Fawcett 8465 (UCWI). 

Unlocalized: 1787, Swartz (BM, type-collection). 

Columnea rutilans is a western Jamaican species (Text-fig. 17) easily recognized 
by the conspicuous red colouring on the under-side of the leaves and the calyx. The 
calyx-segments are, moreover, deeply and pinnately lobed and clothed with long 
reddish hairs. Although occasionally epiphytic, it mostly grows in pockets of humus 
on limestone rocks. Despite its distinctness from other species, it hybridizes with 
C. urbanii when, as at Top Hill, Manchester, it grows at the same place. 



the jamaicanspecies ofcolt/miv£^ and a lloplect us 231 

Hybrids 

Columnea rutilans x urbanii (Text-fig. i6). 

A number of hybrids between ornamental cultivated species of Columnea has been 
raised in gardens, among those named for horticultural purposes being C. X hanksii 
Lynch (C. oerstediana Klotzsch ex Oerstedt X C. schiedeana Schlechtendal), C. 
X euphora H. E. Moore (C. gloriosa Sprague X C. Upidocaula Hanst.), C. X lemoinei 
Garnier (C. glabra Oersted X C. magnifica Oersted) and C. X vedrariensis Mottet 
(C. magnifica x C. schiedeana). Unless, however, plants of two species grow 
together or within the territory of one individual humming-bird, hybridization is 
unlikely often to take place in the wild. Jamaica provides the first recognized 
example of its occurrence. 

In April, 1962, Dr. C. D. Adams collected at Top Hill near Walderston, Manchester 
(Text-fig. 16) a Columnea [Adams 11085; UCWI) which agreed with C. rutilans in 
having its leaves red beneath but diverged in its shorter leaves with longer hairs on 
the upper surface and its toothed rather than laciniate calyx-segments. He surmised 
that it might be a hybrid of C. rutilans with another species. He also collected 
between Contrivance and Top Hill a Columnea {Adams iioyy; UCWI) with leaves 
like C. urbanii but flowers longitudinally red-striped which likewise did not fit any 
recorded species. Earlier, in 1924, C. Norman {Norman 6g; BM) collected near 
Mandeville, Manchester, an anomalous specimen. 

Investigation of the Top Hill area by Dr. Brian D. Morley in 1966 revealed the 
existence here of both C. rutilans and C. urbanii and a series of intermediate plants 
evidently derived from their hybridization. These two species are so distinct in the 
shape, size, colour and pubescence of their leaves and calyx-segments, as well as the 
colour of the corolla, that an isolated intermediate specimen of unknown provenance 
could easily be taken to represent an independent species not closely connected with 
either. Such a specimen, for example, is Morley 334, hybrid 6, having elliptic leaf- 
blades 6-9 cm. long, 3-5-4-5 cm. broad, reddish below, also slightly toothed calyx- 
segments covered with minute bristles and long reddish multicellular hairs, and a 
yellow corolla with similar reddish hairs. The ranges of C. rutilans and C. urbanii 
overlap in Manchester and upper Clarendon and it is within these parishes that the 
plants combining their characters occur, sometimes where neither of the two species 
is now present. This may be due to the disappearance of these species but it may 
also result from the dispersal of seed of hybrid origin by birds and later by self- 
pollination and inbreeding of a colony thus established outside the original locality. 
Morley collected ten nothomorphs at Top Hill in which the characters of C. rutilans 
and C. urbanii were variously mingled. 

The following gatherings also appear to show traces of hybrid origin, although much 
closer to C. urbanii than to C. rutilans: 

Manchester: Grounds of West Indian Training College near Mandeville, 760 m., 
March 1958, Proctor 17455 (BM; IJ). Providence, one and a half miles west-south- 
west of Newport, 800-820 m., Jan., 1961, Proctor 2igo5 (IJ). Rose Hill, 760 m., 
Oct. 1965, Morley 2y4C (BM). Moorlands estate, two miles northeast of Spur Tree, 



232 THE JAMAICAN SPECIES OF COLUMN AE AND ALLOPLECTUS 

760 m., Jan. 1956, Howard, Proctor & Steam 14^01 (IJ). Newport, 680-760111., 
Feb. 1966, Powell & Green 166 j (IJ). 

Clarendon: Peckham Woods, 760 m., March 1956, Robertson 2y2o (UCWI; IJ). 



ALLOPLECTUS Mart. 
Alloplectus Mart., Nov. Gen. & Sp. PI. Brasil. 3 : 53 (1829), nom. cons. 

Shrubs or herbs. Leaves opposite, those of a pair equal to unequal, usually long- 
petiolate. Flowers solitary or fasciculate in leaf -axils or borne in short axillary 
racemes, with bracts. Calyx 5-partite. Corolla inferior, erect or oblique in the 
calyx, gibbous but not spurred at base, the tube cylindric or ventricose, contracted 
or not expanded in the throat, the limb regular or almost so with five short rounded 
lobes. Stamens four, didynamous, the filaments united in the lower part, free and 
filamentous above, the anthers free or cohering in pairs, their loculi parallel, longi- 
tudinally dehiscent. Disc gland large and dorsal, sometimes accompanied by 
smaller glands. Ovary superior; stigma usually 2-lobed. Fruit a capsule or berry; 
seeds minute, longitudinally or obliquely striate. 
Lectotype (see p. 184): A. sparsiflorus Mart. {^ A. hirtellus (Schott) Preston). 

The genus Alloplectus as defined above comprises about 80 species, native to tropical 
America, including the Caribbean islands. 

I. Alloplectus pubescens (Griseb.) Fawcett, Provis. List Flow. PI. Jam. : 28 
(1893). — Urban, Symb. Antill. 2 : 358 (1902). — Morton in Contr. U.S. Nat. 
Herb. 29 : 15 (1944). (Text-figs. 19, 29G-H.) 

Pterygoloma pubescens Griseb., Fl. Brit. W. Ind. Is. : 464 (1862). 
Columnea pubescens (Griseb.) Kuntze, Rev. Gen. PI. 2 : 472 (1892). 

Shrub of dwarf creeping habit, the stems grey, slender, 2-3 mm. thick, hirsute with 
colourless hairs. Leaves paired, those of a pair almost equal to markedly unequal, 
short-stalked, green and densely hirsute on both surfaces, some of the hairs on the 
upper surface glandular, the larger leaf with the blade elliptic or narrowly obovate, 
symmetric, the apex acuminate, the margin inconspicuously denticulate, the base 
cuneate, 4-5-8 cm. long, 2-3 cm. broad, the smaller leaf often half as long as the 
larger leaf; petiole 2-10 mm. long. Flowers solitary in leaf-axils. Pedicel 7-15 mm. 
long, hirsute. Calyx divided to the base, green, hirsute on both surfaces withJ^ 
colourless glandular hairs; segments narrowly elliptic, acuminate, c. i cm. long, entircj 
Corolla tubular, slightly swollen at the middle, c. 2 cm. long, reddish orange, the tube 
c. 4 mm. broad at the middle, strigose outside, the lobes almost equal, 1-2 mmJ 
long. Stamens included; filaments glabrous; anthers c. i-i mm. long. Ovar'^ 
glabrous; style glabrous. Berry pink or purplish, c. 5 mm. broad. 

Portland: 610 m., Wilson (K; GOET, type). John Crow Mountains, one and a| 
third miles south-west of Ecclesdown, 460-610 m., Aug. 1954, Webster 556g (BM; IJ). 
Same locality, Dec. 1955, Proctor 11346 (IJ). Same locality, Jan. 1956, Howard,^ 
Proctor & Steam 14792 (IJ). 



THE JAMAICAN SPECIES OF COLUMN AE AND ALLOPLECTUS 233 

Alloplectus pubescens is a rare and inconspicuous species confined to the rain forest 
of the mountains in the extreme east of Jamaica (Text-fig. 19) and differs from other 
West Indian species of Alloplectus in its entire sepals, which are hirsute on both 
surfaces. The corolla is variously described as " pale orange ", " orange " and 
" pale salmon-pink ", the fruit as " pink " or " light rose-purple ". 




Fig. 29. Alloplectus grisebachianus (Kuntze) Urban: A, flowering shoot [Proctor 22648, 
^6539) ; B, section of flower after anthesis (Harris 333) ; C, calyx (Harris 553) ; D, anthers 
[Harris 555); E, anther, back view [Harris 353); F, gynoecium (Harris 333); Alloplectus 
pubescens (Griseb.) Fawcett: G, calyx [Wilson); H, seed (Webster 536g). Alloplectus 
ambiguus Urban: J, section of ovary (after H. E. Moore, 1957). 



234 THE JAMAICAN SPECIES OF COLUMN AE AND ALLOPLECTUS 

2. Alloplectus grisebachianus (Kuntze) Urban, Symb. Antill. 2 : 357 (1901). — 
Morton in Contr. U.S. Nat. Herb. 29 : 18 (1944). (Text-figs. 18, 29A-F.) 

Pterygoloma cristatum Griseb., Fl. Brit. W. Ind. Is. : 464 (1862). 

Columnea grisehachiana Kuntze, Rev. Gen. PI. 2 : 472 (1892). 

Alloplectus cristatus (Griseb.) Fawcett, Provis. List Flow. PI. Jam. : 28 (1893); non A. 

cristatus (L.) Mart, ex G. Don (1838). 
Alloplectus grisebachianus var. ochrotrichus Urban, Symb. Antill. 5 : 497 (1908). — Morton, 

loc. cit. (1944). 
Columnea hunnewellii L. B. Smith in Rhodora 39 : 275 (1937). 

Shrub of dwarf creeping habit, branched, sometimes forming a bush with 
descending shoots, the stems grey, slender, 2-3 mm. thick, densely hirsute when 
young with reddish hairs or colourless hairs, the leaf-scars prominent. Leaves 
paired, those of a pair mostly unequal, short- or long-stalked, green and hirsute on 
both surfaces with long rather appressed hairs, the larger leaf of the pair with the 
blade elliptic or ovate, slightly or distinctly asymmetric, the apex shortly acuminate, 
the margin shallowly denticulate, the base rounded, 2-5-6 cm. long, i'5-3 cm. 
broad, the smaller leaf sometimes half as long as the larger leaf and falling earlier; 
petiole 3-10 mm. long. Flowers solitary in the leaf-axils. Pedicel 2-10 mm. long, 
hirsute. Calyx divided to the base, green, hirsute outside with long reddish hairs 
or colourless hairs, glabrous inside; segments laciniate, c. 10-12 mm. long, with two 
to five prominent teeth on each side. Corolla tubular, slightly swollen at the middle, 
c. 2 cm. long, dull purplish red, the tube c. 5-6 mm. broad at the middle, hirsute 
outside with long spreading hairs and very minute appressed hairs, the lobes almost 
equal, c. i mm. long. Stamens included; filaments glabrous; anthers c. i mm. long. 
Ovary glabrous; style minutely pubescent in the upper part. Berry white, c. 7 mm. 
broad. 

Trelawny: near Troy, 610 m.. May 1903, Harris 854g (BM; US; lectotype 
collection of Alloplectus grisebachianus var. ochrotrichus). 

Manchester: Olive River, Christiana district, 910 m., Oct. 1901, Harris 8413 
(BM). 

Clarendon: Peckham Woods, 750 m., Dec. 1917, Harris I2j66 (K; US). Same 
locality, Dec. 1955, Steam 10 (BM). 

St. Ann: Albion, March 1850, Purdie 555 (K, lectotype of Pterygoloma cristatum). 
Dry Harbour Mountains, two miles west of Albion, 600 m., Jan. 1956, Steam 1^8 
(BM; K). Mount Diablo, Feb. 1916, Ridley (K). Same locality, Jan. 1938, 
Hunnewell (BM; K; US). Union Hill, near Moneague, 750 m., Britton & Hollick 2^44. 
(NY). Hollymount, Feb. 1905, Harris 88gi (NY). Same locality, Sept. 1906, 
Britton yi8 (NY). Along the road to Hollymount, Mount Diablo, 800 m., Aug. 1957, 
Proctor i653g (BM). Ramble, Claremont, 450 m., Jan. 1898, Fawcett & Harris yi68 
(BM). 

Unlocalized: Alexander (GOET). Wullschldgel (GOET) 

The transfer by Fawcett in 1893 of Grisebach's epithet cristatus from Pterygoloma 
to Alloplectus created a later homonym of Alloplectus cristatus (L.) Mart, ex G. Don, 
based on Besleria cristata L. (1753). In 1892, however, Kuntze united Alloplectus, 
Nematanthus and Hypocyrta with Columnea and as he could not use Grisebach's 



I 



THE JAMAICAN SPECIES OF COLUMNAR AND ALLOPLECTUS 235 

epithet owing to the prior claim of Linnaeus's pubhshed under Besleria, which 
resulted in the name Columnea cristata (L.) Kuntze, he coined the new name Columnea 
grisebachiana for Pterygoloma cristatum Griseb. Thereby he provided a legitimate 
epithet for use under Alloplectus. For a biography and bibliography of August 
Heinrich Rudolf Grisebach (1814-79), the Gottingen professor of botany whom the 
name commemorates, see Journ. Arnold Arb. 46 : 245-260 (1965). 

Alloplectus grisebachianus is restricted to the central uplands of Jamaica (Text-fig. 
18). The typical and commonest form of the species is notable for the long reddish 
multicellular hairs which clothe the branchlets, petioles, leaf-blades, pedicels, calyx 
and corolla. To this belongs the type-specimen of Columnea hunnewellii L. B. Smith 
collected by Hunnewell and Griscom in March 1936, on Mount Diablo. In speci- 
mens collected by Harris at Troy {Harris 8^4g) and at Olive River {Harris 841 j) the 
hairs are colourless and Urban distinguished them as A. grisebachianus var. ochro- 
trichus in 1908. A gathering made by Hunnewell in January 1938 on Mount Diablo 
includes, however, both forms; the loss of colour in the hairs has probably taken 
place independently in several localities. 

Alloplectus ambiguus Urban is a species not of Jamaica but of Puerto Rico; because 
it is adequately known from wild and cultivated material and is allied to the Jamaican 
species, it has been included above in the taxonometric survey (pp. 207-214) to 
increase the representation of Alloplectus and in Fig 29 (p. 233) to illustrate the 
placentation of these plants. 



SPECIAL LITERATURE 

Bond, J. 1947. Field Guide to Birds of the West Indies. Ed. 2. New York. 

British Museum (Natural History). 1966. Report on the British Museum {Natural History), 

ig63-ig63. London. 
Burman, J. 1756. Plantarum Americanarum Fasciculus Quartus. Amsterdam & Leiden. 
Eberle, p. 1956. Cytologische Untersuchungen an Gesneriaceen. Chromosoma 8 : 285-316. 
Faegri, K. & VAN der Pijl, L. 1966. The Principles of Pollination Ecology. Oxford. 
Fawcett, W. 1893. A Provisional List of the Indigenous and Naturalised Flowering Plants of 

Jamaica. Kingston, Jamaica. 
Fussell, C. p. 1958. Chromosome numbers in the Gesneriaceae. Baileya 6 : 117-125. 
Gower, J. C. 1966. Some distance properties of latent root and vector methods used in 

multivariate analysis. Biometrika 53 : 325-338. 
Grant, V. & Grant, K. A. 1966. Records of hummingbird pollination in the western American 

flora. Aliso 6 : 51-66. 
Greenewalt, C. H. i960. Hummingbirds. New York. 

Grisebach, A. H. R. 1859-64. Flora of the British West Indian Islands. London. 
Harborne, J. 1966a. Ca.fieic acid ester distribution in higher plants. Z. Naturforsch. 

21B : 604-605. 
1966&. Comparative biochemistry of the fiavonoids — II. 3-desoxyanthocyanins and 

their systematic distribution in ferns and gesnerads. Phytochemistry 5 : 589-600. 
Kendrick, L. B. & Weresub, L. K. 1966. Attempting neo-Adansonian computer taxonomy 

at the ordinal level in the Basidiomycetes. Syst. Zool. 15 : 307-329. 



0^^c 



'<S.y 



236 THE JAMAICAN SPECIES OF COLUMN AE AND ALLOPLECTUS 

Lee, R. E. 1962a. Chromosome numbers in the Gesneriaceae. Baileya 10 : 33-45. 

1 9626. Achimenes coccinea, a tetraploid; and additional counts of chromosome numbers 

in the Gesneriaceae. Baileya 10 : 146. 

1964. Additional chromosome numbers in the Gesneriaceae. Baileya 12 : 159. 

1966. Additional chromosome numbers in the Gesneriaceae. Baileya 14 : 35-36. 

1967. Additional chromosome numbers in the Gesneriaceae. Baileya 15 : 118. 

& Grear, J. W., Jr. 1963. Additional chromosome numbers in the Gesneriaceae. 

Baileya 11 : 131. 
Leeuwenberg, a. J. M. 1958. The Gesneriaceae of Guiana. Acta bot. neerl. 7 : 291—444. 
MooRE, H. E. 1957. African Violets, Gloxinias, and their relatives. New York. 
MoRLEY, B. D. 1966. Columnea and aspects of its evolution. Sci. Notes <&- News, Jamaica 

2 : 13-15- 

1967. In A. Love, lOPB chromosome number reports. Taxon 16 : 215-222. | 

Morton, C. V. i944fl. The West Indian species of Columnea. Contr. U.S. natn. Herb. 

29 : 1-15. 

1944&. The West Indian species of Alloplectus. Contr. U.S. natn. Herb. 29 : 15-19. 

1963. Another Columnea new to cultivation. Baileya 11 : 23-27. 

PiTELKA, F. A. 1942. Territoriality and related problems of North American hummingbirds. 

Condor 44 : 187-204. 
Raymond, M. 1965. Species of Columnea and Trichantha in cultivation in the Montreal Botanic 

Garden. Mem. Jard. bot. Montreal 57. 
Rogers, O. M. 1954. Chromosome counts in Gesneriaceae. Baileya 2 : 14-18. 
Sealy, J. R. 1950. Nomocharis and Lilium. Kew Bull. 1950 : 273-297. 
Sheals, J. G. 1965. The application of computer techniques to Acarine taxonomy. Proc. 

Linn. Soc. Lond. 176 : 11-21. 
Sims, R. W. 1966. The classification of the Megascolecoid earthworms: an investigation of 

Oligochaete systematics by computer techniques. Proc. Linn. Soc. Lond. 177 : 125-141. 
SoKAL, R. R. & Sneath, p. H. A. 1963. Principles of Numerical Taxonomy. San Francisco 

& London. 
Stearn, W. T. 1959. A botanist's random impressions of Jamaica. Proc. Linn. Soc. Lond. 

170 : 134-147. 

1968. Observations on a computer-aided survey of the Jamaican species of Columnea 

and Alloplectus. In V. Heywood, Modern Methods in Plant Taxonomy: 219-224. 
SwARTZ, O. 1788. Nova Genera et Species Plantarum seu Prodromus. Stockholm, Uppsala 

& Abo. 
Urban, I. 1901. Enumeratio Gesneriacearum. Symb. Antill. 2 : 344-388. 

1909. Nova genera et species IV. Symb. Antill. 6 : 1-55. 

Watson, L., Williams, W. T. & Lance, G. N. 1967. A mixed-data numerical approach to 

angiosperm taxonomy: the classification of Ericales. Proc. Linn. Soc. Lond. 178 : 25-35. 



W. T. Stearn, D.Sc, Sc.D. 

Dept. of Botany 

British Museum (Natural History) 

Cromwell Road 

London, S.W.7 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 14 



MrVK^'M >KAEK«At. H«»rO«Vi^ 




,1,, Io»»iiiMr ax, i955 

r VO(»I*a liM«tOE« hiii«ia». 
frail lii« •piphyt*! s»iyjt r»ai c 
>«XIow, str*Kj£*d with orange. 
,., S»org« a. Proctor s„ iiajZ 



aivuiK riisi^ 



Columnea jamaicensis Urban {Proctor 11 232 ; representative specimen in 
Herb. Brit. Mus. (Nat. Hist.), London). 
EOT. 4. 5. ^2 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 15 



CO 




Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 16 




urhaiiii 






Columnea urbanii Stearn [Steam 361; holotype in Herb, Brit, Mus. (Nat. Hist.), London) 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 17 




Columnea proctorii Stearn [Steam 451; holotype in Herb. Brit. Mus. (Nat. Hist.), London). 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 18 







proctor I 



Cohimnea proctor a Steam [Steam 4^1; isotype in Herb. Brit. Mus. (Nat. Hist.), London). 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 19 




UNlTtO »TATW M*TIQNAL MUSEUM 
PLANTS OF JAMAICA 



^^ /i< i^^ /Zffjf 



Columnea subcordata Morton {Miller 1556; holotype in U.S. Nat. Herb., Washington) 



\ 



Bull. Br. Mus. nai. Hist. (Bot.) 4, 5 



PLATE 20 




argfntea 




FLORA OF JAMAICA 

M»i»i li»«w»i««t(H' HIT 2»oo ft. D«Tt as. i;. i»b4 

i.«*on SwaiMrtli, apKH( roekt b«hln<l tJi« MorcvtBi 
etank i« $«p «f htli. 

•MWit. cr»|', i« to o>u long, 1m»»« .pr»,4iB 
•11^, «!• tsBt*«t 18 U t>«. limj. S a>. fcro«4t 
(wljf* U#t 9-nni ooratU ptit ]r*ila«i tnilt 
»llip*ali, ;Uk, 11 w. !<«(> S mi. t>n>*4. 



Cohminea argentea Griseb. {Steam j^^g; topotype in Herb. Brit. Mus. (Nat. Hist.), London). 



\ 



^ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 5 



PLATE 21 




rut i kins 



Columnea rutilans Swartz [Steam 1003; representative specimen in Herb. Brit. Mus. (Nat. 

Hist.), London). 



\ 







::"t:v: 



' ■'.',.>/ 



.i^ '^< "•^^ "^;:'-':v-^ •:^:;)I- ■''■?;'-; i<N'i; ,^^^^^^^ 








PkiNTJKD IT* GREAT BftlTAIBr 
BY ADLARD & SON LIMITED 
BARTHOLOMEW PRESS, DORKIKG 



NEW OR LITTLE KNOWN 

HIMALAYAN SPECIES OF 

SWERTIA AND VERATRILLA 

(GENTIANACEAE) 




H. SMITH 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 6 

LONDON : 1970 



NEW OR LITTLE KNOWN HIMALAYAN 

SPECIES OF SWERTIA AND VERATRILLA 

(GENTIANACEAE) 



I 




fl 



BY 

HARRY SMITH )L-it^X 

Institute of Systematic Botany, University of Uppsala 



pp. 237-258 ; 7 Text-figures, Plates 22-37 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 6 

LONDON: 1970 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is issued 
in five series corresponding to the Departments of the 
Museum, and an Historical series. 

Parts will appear at irregular intervals as they 
become ready. Volumes will contain about three or 
four hundred pages, and will not necessarily be com- 
pleted within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4 No. 6 of the Botanical series. 



World List abbreviation 
Bull. Br. Mus. not. Hist. (Bot.). 



Trustees of the British Museum (Natural History), 1970 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 



Issued 18 August, 1970 



Price £1 lis. 
(£i-55) 



NEW OR LITTLE KNOWN HIMALAYAN 

SPECIES OF SWERTIA AND VERATRILLA 

GENTIANACEAE 

By HARRY SMITH 

SUMMARY 

Ten species of Swertia : S. acaulis, S. grandiflora, S. virescens, S. pseudo-hookeri, S. 
staintonii, S. assamensis, S. crossoloma, S. splendens, S. gracilescens and 5. burmanica 
are described as new. The first eight of these, with 5. hookeri C. B, Clarke and 
5. candelabrum H, Smith, form a distinct group within Swertia sect. Swertia and a 
key is provided for this group. The new name 5. franchetiana is proposed for 5. 
striata Franch., a later homonym of 5. striata CoUett & Hemsl. Kingdonwardia 
codonopsidoides Marq. is shown to be a synonym of Swertia racemosa (Griseb.) C. B. 
Clarke and reasons are given for not maintaining Kingdonwardia as a genus distinct 
from Swertia. Extensions of range are recorded for S. kiharae Kitam. and the 
previously described species mentioned above. 

Certain errors in Franchet's original account of Veratrilla are pointed out, Swertia 
mekongensis Balf. f. & Forr. is treated as a synonym of Veratrilla baillonii Franch. 
and Swertia burkilliana W. W. Smith is transferred to Veratrilla. Extensions of 
range of both species are recorded. 

INTRODUCTION 

Until this century the world's herbaria contained few specimens of plants from 
the Himalayan region. In the last few decades, however, a very considerable 
amount of material has been collected in Tibet, Assam, Bhutan, Sikkim, and, more 
recently, Nepal, much of this in the course of expeditions under the auspices of the 
British Museum (Natural History). This paper deals with those specimens of the 
genera Swertia and Veratrilla that have been collected in those countries and that 
either represent new species or provide new information about species previously 
described. Except for a few specimens from the Royal Botanic Garden, Edinburgh, 
indicated by " (E) ", all those cited are in the herbarium of the British Museum 
(Natural History) ; there are duplicates of the great majority in Uppsala. 



240 HIMALAYAN SWERTIA AND VERATRILLA 

SWERTIA L 

SwERTiA L., Sp. PI. I : 226 (1753) ; Gen. PI., ed. 5 : 107 (1754). 

Sectio SWERTIA 

Swertia subgen. Euswertia C. B. Clarke in Hook, f., Fl. Brit. Ind. 4 : 127 (1883), nom. invalid. 

Swertia sect. Euswertia Gilg in Engl. & Prantl, Naturl. Pflanzenfam. 4 (2) : 88 (1895), nom. 

invalid. 

Swertia kiharae Kitam. in Acta Phytotax. & Geobot. 16 : 133 (1956). (Plate 22 ; 
Text-fig. I.) 

Chitral : Gangallat Gol, S.W. of Chitral, 2,700 m., amongst shrubs at edge of 
stream, petals and filaments white, anthers black, 18 June 1958, Stainton 2yo8. 
Golan Gol, 3,000 m., edge of stream, white, 13 July 1958, Bowes-Lyon 63. 






Fig. I. Swertia kiharae Kitam. : a, corolla lobe with nectary and stamen 
b, calyx ; c, ripening capsule {Stainton 2yo8). All X5. 



These are interesting finds of a species originally described from Afghanistan. No 
species of the Swertia aucheri Boiss. group has been recorded until now from the 
Indian sub-continent. 

The Swertia hookeri group of species 

The botanical explorations in the remote parts of the Himalayas during this 
century have shown that Swertia hookeri C. B. Clarke, originally described from a 
single collection from Sikkim, is not an isolated species but is instead a member of a 
complex group consisting of not less than ten distinct species, all obviously of close 
affinity. Unfortunately the material in herbaria is not rich ; collectors do not rehsh 
putting such robust and unwieldy plants in the press. 



HIMALAYAN SWERTIA AND VERATRILLA 241 

This group is well differentiated from other members of the section Swertia by the 
following characters : 4-merous fairly large to very large flowers, single nectary, 
robust growth, thick hollow stem, which can be o-8 m. high or more (the abbreviated 
5. acaulis H. Smith excepted), and seeds globose and unwinged or much flattened 
and winged. 

A distinctive character of unusual nature divides this group into two parts, each 
with five species. In the first part the margin of the leaves and sepals is glabrous, 
though a cellular papillosity is visible at a high magnification (text-fig. 2, b). In 
the other the margin is cut into hair-like, chlorophyllaceous lacinules only 0-3-0 -5 
mm. long (text-fig. 4, b, c). No specimen has been found that deviates in the 
slightest way from one or other of these two patterns. The species development 
seems to have followed parallel lines in these two groups. In both we have species 
with naked or with fimbriate nectaries, and species with cymose or with pyramidal 
inflorescences. Whether there is a similar parallelism in types of seed cannot yet 
be determined ; there are four species of which there is not available ripe seed nor 
ovules so far advanced that the shape of the seed can be safely presumed. At present 
we know that globose wingless seeds are present in both groups but flattened winged 
seeds are so far known only in the first. 

The following is a key to this Group of species : 

Margin of leaves and sepals entire, with obsolete cellular papillae (text-fig. 2, b) : 
Nectary naked : 

Inflorescence cymose, many-flowered ; seed flattened : 

Plant 20-80 cm. tall ; inflorescence of distant axillary cymes, rarely 

moderately branched ; seed discoid, circularly winged . 5. hookeri 

Plant subacaulous, stem at most 2 cm. tall ; seed oval, the obtuse ends 

unequally and irregularly winged . . . . S. acaulis 

Inflorescence pyramidal ; seed globose, not winged : 

Large robust plant 60 cm. or more tall ; petals reddish, 3 cm. long by i-6 
cm. broad ; inside base of sepals with a ring of black hairs 

S. grandiflora 

t Plant 20-30 cm. tall ; petals greenish, i •2-2-0 cm. long by 0-7-1-2 cm. 

broad ; inside base of sepals glabrous . . . .5. virescens 

Nectary densely fimbriate ; plant 30-40 cm. tall ; inflorescence pyramidal ; 
flowers yellowish-greenish, with glabrous and scarcely winged pedicels; 
seeds globose, not winged . . . . . S. pseudo-hookeri 

Margin of leaves and sepals minutely and somewhat irregularly lacinulate, lacinules 
chlorophyllaceous, hair-like, 0-3-0 -5 mm. long (text-fig. 4, b, c) : 
Nectary naked, its margin smooth : 

Stem deeply 5-grooved ; inflorescence pyramidal, copiously branching from 
base and nodes ; wings of stem and pedicels ciliate ; petals 2 cm. long 
by 1-7 cm. broad ; seed globose, not winged . . . S. staintonii 

Stem terete or nearly so ; inflorescence of axillary cymes not or moderately 
branched ; wings of stem and pedicels glabrous ; petals 2-5 cm. long by 
2 cm. broad ; (seed not seen) . . . . . S. assamensis 

Nectary toothed or fimbriate at the margin : 



242 HIMALAYAN SWERTIA AND VERATRILLA 

Margin of nectary toothed with pointed projections not more than 0-5 mm. 

long, less than i/3rd of the diameter of the nectary ; flowers few on robust, 

curvingly ascendent branches . . . . 5. candelabrum 

Margin of nectary fimbriate with filamentous projections 1-2 mm. long, ^ or 

more of the diameter of the nectary ; inflorescence pyramidal : 

Connective of anthers prolonged, often claw-like ; flowers pale creamy 

green ; petals i-g cm. long by i-o cm. broad ; wings of pedicels ciliate 

5. crossoloma 

Connective of anthers not prolonged ; flowers brick red ; petals 2-5 cm. 

long by 2-0 cm, broad ; wings of pedicels glabrous except at point of 

junction with the decurrent sepals . . .5. spiendens 

SwERTiA HOOKERi C. B. Clarke in Hook, f., Fl. Brit. Ind. 4 : 127 (1883). (Plate 23 
Text-fig. 2 a-d.) 

Tibet : Kongbo, Doshong La, 4,000 m., on grassy slope at base of precipice, herb 
to I m., calyx green, corolla greenish yellow with faint browny purple veins, filaments 
and ovary purple-black, anthers dull yellow, 15 July 1938, Ludlow, Sherriff and 
Taylor 5256. Same locality, amongst shrubbery, calyx green, corolla yellowish green, 
anthers and ovary dark green, 16 Aug. 1947, Ludlow, Sherriff and Elliot 14376. 

Nepal : Barun Khola, N. of Num, Arun Valley, 3,800 m., on open slopes, corolk 
greenish brown, filaments white, 12 June 1956, Stainton 641. S. of Topke Gola, 
Arun-Tamur watershed, 4,300 m., on open slopes, corolla, filaments and anthers 
bronze-coloured, 9 July 1956, Stainton goj. 

SiKKiM : Changu, 3,700 m., in turfy level space on hill, red flowers, spikes 60-90 
cm., 8 Sept. 1913, R. E. Cooper 832. North Chakung Chu, 4,000 m., on peat hillside, 
I m. tall, red flowers, many whorls, 20 Sept. 1913, R. E. Cooper gi8. 

Bhutan : Ju La Mangde Chu, 4,300 m., in open alpine zone, calyx green, corolla 
brownish red, 19 July 1949, Ludlow, Sherriff & Hicks i68g2. Parshong, 4,ooo-3,8o( 
m., tall, red, 27 July 1914, R. E. Cooper 1993. 

Clarke's original description, although based on only two gatherings, is very 
accurate and complete and there can be no doubt about the identity of his plant. 



Swertia acaulis H. Smith, sp. nov. (Plate 24 ; Text-fig. 2e.) 

Swertia acaulis H. Smith ex Nilsson in Grana Palyn. 7:111 (1967), nom. nud. 

Omnino glabra, radix validus verticalis, caulis valde abbreviatus 1-2 cm. altusi 
flores numerosos, cymosos, subaequaliter 3-4 cm. longe pedicellatos edens. Habitu 
5. multicaulis D. Don^ subsimilis sed distat : floribus multo majoribus albis (nee 
caeruleis), nectario perfecte nudo (nee fimbriato), seminibus complanatis, ambitu 
ovalibus, in apicibus inaequaliter alatis (nee subglobosis exalatis). 

Folia rosularia desunt, caulina aggregata, numerosa, lanceolata, acuta, trinervia, 

^ C. B. Clarke in Hook, f., Fl. Brit. Ind. 4 : 129 (1883) says that 5. multicaulis is 5-merous, whereas it 
is 4-merous, as correctly stated in the original diagnosis by D. Don, Prodr. Fl. Nepal : 128 (1825). 



HIMALAYAN SWERTIA AND VERATRILLA 



243 



5-7 cm. longa, ad 0-9 cm. lata, lamina in petiolum subaequilongum sensim angustata. 
Flores 4-meri ; calyx fere ad basin fissus, lobi ovati, acuti, 11-16 mm. longi et 5-7 mm. 
lati ; coroUae tubus perbrevis, lobi obovati, rotundato-truncati, apice minute 
lacerulati, 18 mm. longi et 11 mm. lati, nectario nudo nigro ; stamina tubo adnata, 
filamentis liberis 11 mm. et thecis 2 mm. longis, connectivo antherarum fere 0-5 mm. 
longe aristulatim protracto ; ovarium anguste ovatum in stigmas sessiles angustatum ; 
semina complanata, ambitu ovalia, i-oxo-6 mm. magna, in apicibus valde inaequa- 
liter alata. 




N 



I 




CO 



Fig. 2. Swertia hookeri C. B. Clarke : a, corolla lobe with nectary and stamen ; b, margin 
of leaf, with chlorophyllaceous area dotted ; c, middle part of rosular petiole {Ludlow, 
Sherrijf &> Elliot 14326) ; d, seed from above and in cross section (holotype) Swertia 
acaulis H. Smith : e, seed from above [Lull Dhowj ozgi). b X 50, e x 10, a, c, d X5. 



Nepal : Foot of Rolwaling, 3,700-5,500 m., yellow, 1930, Loll Dhowj 02gi (holo- 
type). Dudkund, 5,500 m., yellow, 24 Aug. 1932, K. N. Sharma E504. Wabak 
Khola, east of Num, Arun Valley, 4,600 m., on open slopes, petals white, filaments 
and anthers blue, 12 Aug. 1956, Stainton 1334. 



244 HIMALAYAN SWERTIA AND VERATRILLA 

Swertia grandiflora H. Smith, sp. nov. (Plate 25.) 

Planta 60 cm, vel ultra alta S. candelabra H. Smith subsimilis, sed distat : margine 
foliorum et sepalorum integro (nee lacinulato), floribus majoribus 3 cm. longis, 
calycis lobis late ovalibus obtusis (nee cordato-triangularibus, acutis). 

Caulis robustus ad i-6 cm, diametiens, ut rami et pedicelli modeste quadrialatus, 
glaber, Inflorescentia pyramidalis, ramis robustis, paucifloris e nodis 2 arcuatim 
adscendentibus. Folia margine Integra, rosularia ad 18 cm. longa et 6 cm. lata, 
7-13 nervia, lamina elliptica apice acuta in petiolum latum brevem vel longiorem 
angustata, caulina perpauca ad 9-5 cm, longa et 6 cm. lata, sessilia, nervis c. 9 Flores 
patuli ad 5-5 cm. diametientes, rubroviride striati ; calycis tubus perbrevis basi 
interne breviter nigro-barbatus, lobi inaequales, duo majores ad 1-9 cm. longi et 
1-5 cm, lati, ovales, obtusi vel subobtusi, glabri ; corollae tubus brevis, lobi oblongi, 
apice rotundati, ad 3 cm. longi, superiori parte I-5-I-7 cm, lati, nectario nudo sub- 
quadrangulari margine pauUo elevato cincto ; stamina tubo adnata, filamentis 
liberis crassis 10 mm, longis, antheris caeruleis 3 mm, longis ; ovarium conicum 
stigmatibus parvis sessilibus coronatum ; capsula matura (e Ludlow & Sherriff 1040) 
2 cm. longa, valvis apice recurvantibus, pedicellis ad 4-10 cm. prolongatis ; semina 
ovali-globosa, exalata, i7-2-o — i •4-1-6 mm, magna, testa levi, 

Bhutan : Shingbe, Me La, 4,000 m., on open hillside, very common, flowers stripec 
dull red and green, 14 June 1949, Ludlow, Sherriff & Hicks 20^32 (holotype). Me La] 
4,300 m,, open grassy hillside above tree line, 6 Oct. 1934, Ludlow & Sherriff 1040. 

Swertia virescens H. Smith, sp. nov. (Plate 26 ; Text-fig, 3c, d.) 

Swertia virescens H. Smith ex Nilsson in Grana Palyn. 7 : 112 (1967), nom. nud. 

Species 15-20 cm. alta, S. hookeri C. B. Clarke affinis sed distat : planta multc 
minore, inflorescentia pyramidale, floribus vix 1-5 cm, longis, foliis omnibus obtusis| 
(nee subacutis) caulinis sessilibus (nee longe petiolatis). 

Caulis glaber, quadrialatus, superiori dimidia parte 3-nodosus, e nodo infimo rami 
4, pluriflori, suberecti, florendi tempore 4-6 cm, longi editi, in nodo supremo floresj 
numerosi cymosi, omnes graciliter pedicellati, pedicellis glabris modeste quadrialatis.| 
Folia rosularia numerosa, lamina ovali, obtusa, 7-nervia, margine Integra, c. 5 cm. 
longa et 2-5 cm. lata, in petiolum ad 1-5 cm, longum et i cm, latum, margine densius 
ciliato, attenuata, folia caulina ovato-ovalia sessilia, Flores virescentes, brunneo- 
striati, nutantes ; calycis tubus perbrevis, lobi ovati, acuti vel subobtusi, ad 10 mmJ 
longi et 5-7 mm. lati, margine integri ; corollae tubus vix 2 mm. longus, lobi cuneato-l 
obovati, apice rotundati et minutissime erosulati, ad 12 mm, longi et 5-7 mm, lati, 
nectario nudo subrotundato ; stamina tubo 1-5 mm. alte adnata, filamentis liberis 
5 mm. longis, antheris 2-3 x 17 mm, magnis ; ovarium anguste ovatum, stigmatibus 
subsessilibus orbicularibus ; capsula, semina non visa, ovulis semievolutis, ut 
videtur, subglobosis exalatis. 

Bhutan : Tsampa (Tolegang), Pangotang, 4,400 m, on open hillside, locally 
common, corolla green striated red brown, 3 July 1949, Ludlow, Sherriff & Hicks 
19323 (holotype). 



HIMALAYAN SWERTIA AND VERATRILLA 



245 



The following two collections are closely allied to Swertia virescens : Tibet, Chick- 
char, Tsari, 4,600 m., on open grassy hillside above rhododendrons, pale cream green, 
filaments dull green, anthers dark brown, stigma pale green, 13 June 1936, Ludlow 
6- Sherriff 2i4g ; Bhutan, Tsampa, Pangotang, 4,300 m., in open wet stony place, 





O O 



Fig. 3. Swertia pseudo-hookeri H. Smith : a, corolla lobe with nectary and stamen ; b, seed 
{Ludlow, Sherriff &■ Hicks ig4i4). Swertia virescens H. Smith : c, corolla lobe with nec- 
tary and stamen ; d, middle part of rosular petiole {Ludlow, Sherriff cS- Hicks, igjzj). 

I All X5. 



corolla white with striations inside, 16 June 1949, Ludlow, Sherriff & Hicks igi68. 
These are both low-growing but very robust plants that are just beginning to flower ; 
their flowers are bigger than those of Swertia virescens and differ in colour, and the 



246 HIMALAYAN SWERTIA AND VERATRILLA 

striation of the petals is hardly noticeable. They probably represent an undescribed 
species which cannot be published as new on the unsatisfactory material yet available. 

Swertia pseudo-hookeri H. Smith, sp. nov. (Plate 27 ; Text-fig. 3a-b.) 
Swertia pseudo-hookeri H. Smith ex Nilsson, Grana Palyn. 7 : in (1967), nom. nud. 

Planta 30-40 cm. alta, caulis teres, inflorescentia laxa e nodis 3-4 anguste pyra- 
midalis. Habitu 5. virescenti H. Smith subsimilis sed nectario longe nigro-fimbriato 
(nee nudo) inter alia distincta. 

Folia rosularia ad 12 cm. longa, lamina elliptica subacuta c. 5 cm. longa et 1-5 cm. 
lata, margine Integra, in petiolum latum margine glabrum laminam longiorem vel 
breviorem sensim attenuata, folia caulina sessilia, sursum decrescentia, in nodis paria. 
Flores albo-virides marginem versus purpureo-striati, 1-5 cm. longe pedicellati, 
pedicellis glabris obsolete alatis ; calycis tubus brevissimus, lobi subaequimagni ovati 
obtusi, c. 10 mm. longi et 5-7 mm. lati, margine integri ; coroUae tubus perbrevis, 
lobi obcuneati, apice truncati et parum erosulati, 1-5 cm. longi et 0-9 cm. lati, 
nectario ovali c. 2 mm. longe nigro-fimbriato, basi non clauso ; stamina tubo adnata, 
filamentis liberis 8 mm. longis et thecis 3-0x1 -6 mm, magnis ; ovarium ovatum, 
stigmatibus parvis subsessilibus ; capsula (e Ludlow, Sherriff & Hicks igSoi) angusta, 
1-5 cm, longa, carpellis apice excurvantibus, seminibus ovato-globosis exalatis c, 
1-4 X 1-2 mm. magnis. 

Bhutan : Tsampa, Marlung, 4,400 m., on open grassy hillside, corolla pale green, 
finely striated purple, 12 July 1949, Ludlow, Sherriff & Hicks 19414 (holotype). 
Upper Mangde Chu, Saga La, 4,700 m., 14 July 1949, ravines in alpine zone, herb 
60 cm. or more in height, calyx pale green, corolla greenish white finely striated 
purple at edges of segments, filaments purplish, anthers black, ovary and style 
greenish blue, 14 July 1949, Ludlow, Sherriff & Hicks i68iy. Mangde Chu, Nam- 
dating, 4,300 m., on steep open hillside, flowers over, 3 Oct. 1949, Ludlow, Sherriff & 
Hicks igSoi. Pumthang, Tibdeh La, 4,000 m., on peaty slopes, showy pyramid of 
white and pale blue flowers, i Sept, 1915, R. E. Cooler 4y6g. 

Swertia staintonii H. Smith, sp. nov. (Plate 28 ; Text-fig. 4.) 

Swertia staintonii H. Smith ex Nilsson in Grana Palyn. 7 : 112 (1967), nom. nud. 

Planta c. 30 cm. alta, inflorescentia pyramidalis e basi et nodis 3-4 ramosa, ramis 
et ramulis suberectis. Species 5. crossolomae H. Smith subsimilis sed distat : caule, 
praesertim inferiori parte, alte 5-sulcato (nee subtereti), floribus majoribus, nectario 
nudo (nee fimbriato), connectivo antherarum subplano (nee protracto, saepe unci- 
nato) . 

Caulis, rami et pedicelli ciliato-alati, alls caulinis in sulcis nee in angulis locatis. 
Folia rosularia numerosa, subacuta, 6-10 cm. longa et 1-2 cm. lata, lamina in 
margine dense lacinulata, lacinulis robuste piliformibus 0-3-0-4 mm. longis, in 
petiolum subaequilongum sensim angustata, petiolo in margine laxe ciliato ; folia 
caulina infimi nodi 5-verticillata, subacuta, lanceolata, c. 7 x 2 cm. magna, breviter 
petiolata. Flores rosei (e coUectore) 1-2 cm. longe pedicellati ; calycis tubus 



HIMALAYAN SWERTIA AND VERATRILLA 



247 



perbrevis, lobi ovato-triangulares, acuti, dense lacinulati, 1-5 cm. longi et i cm. lati, 
marginibus basalibus in alas pedicelli decurrentibus ; corollae tubus vix 2 mm, 
longus, lobi obcuneati margine fere integri, ad 2 cm. longi et 1-7 cm. lati, nectario 
nudo basi rotundato apice subacuto ; stamina tubo 1-5 mm. alte adnata, filamentis 
liberis I cm. longis, antheris 3x2 mm. magnis minutissime apiculatis ; ovarium 
anguste ovatum, stigmatibus orbicularibus subsessilibus ; semina (ex Cooper 1994 
et 2ggo) subglobosa, exalata, i-5-i'9x i-2-i-4 mm. magna. 





^^JA/^'^'i'^'^'^'^''^'^W^tMO\;^yi 



'MAam, 



'W^ 



''^vV, 




I 



Fig. 4. Swertia staintonii H. Smith : a, corolla lobe with nectary and stamen ; b, margin 
of leaf ; c, the same, with chlorophyllaceous area dotted {Stainton 755) ; d, cross section 
of lower part of stem {Cooper 1994). a, b X 5, c X 50, d x 2^. 



Nepal : Arun Valley, Chhoyang Khola, west of Num, 3,800 m., on open slopes, 
corolla pink, filaments black, anthers white, 23 June 1956, Stainton y^^ (holotype). 

Bhutan : Parshong Timpu, 4,000 m., on peaty gravel, common in this one place, 
flowers white, 27 July 1914, R. E. Cooper igg4(E) . Same loc, in alpine turf, pyramid 
of white flowers, 25 Sept. 1914, R. E. Cooper 2ggo{E). 

I have no doubt that Cooper's two gatherings belong to the same species as 
Stainton's. All three have exactly the same unusual morphological organization. 



248 HIMALAYAN SWERTIA AND VERATRILLA 

The difference in colour is of little account. The Cooper specimens are both in a far 
advanced fruiting condition ; the flowers are shrivelled, the leaves withered or gone, 
the capsules mostly empty, but the important characteristics are still there. The 
peculiar arrangement with the deeply groved stem and the 5-verticillate lower 
cauline leaves is exactly the same as in the Stainton plant. So are also the shape 
and margin of the leaves and sepals and the size and form of the petals and nectaries. 
Cooper's notes on the labels " pyramid of white flowers " and " common in this 
one place " show that he was well acquainted with the plant. It is certainly to be 
supposed that he also collected flowering specimens, but no such specimen ever 
reached Edinburgh. Possibly some packages of his got lost on their way during 
the troubled year of 1915. 



Swertia assamensis H. Smith, sp. nov. (Plate 29.) 

Planta robusta c. 30 cm. alta, habitu 5. hookeri C. B. Clarke subsimilis, sed inter 
alia distat : floribus majoribus, marginibus foliorum sepalorumque dense lacinulatis 
(nee integris). 

Caulis glaber modeste alatus. Folia omnia et sepala lacinulata, lacinulis pili- 
formibus vix 0-5 mm, longis ; folia rosularia ad 9 cm. longa, lamina elliptica sub- 
obtusa 2-2-5 cm. lata, in petiolum subaequilongum sensim attenuata, petiolo 
margine sparse ciliata, Flores in cymis 3 distantibus, raro breviter ramosis, dispositi, 
subpatuli, magni, c. i cm. longe pedicellati, pedicellis quadrialatis glabris ; calycis 
lobi duo majores triangulari-ovati ad 13 mm. longi et basi 12 mm. lati, duo minores 
angustiores et breviores ; petala late obcuneata ad 2-5 cm, longa et apicem versus 
2 cm, lata, margine minutissime lacerulata, nectario singulo, nudo, margine basali 
membranaceo modice elevato ; stamina tubo brevissimo adnata, filamentis liberis 
10 mm, longis, thecis robustis 4 mm, longis ; stigmata subsessilia ; capsula, semina 
non visa, 

Assam : Poshing La, 4,000 m,, on cliff ledges and turf slopes, " (= 13797) "' 
22 July 1938, Kingdon-Ward 14008 (holotype). Bhutan frontier, Orka La, 3,400- 
4,000 m,, abundant in alpine pastures where cattle feed, comes up in colonies where 
scrub is cleared, height 25-35 cm,, flowers yellow, streaked red, 7 June 1938, 
Kingdon-Ward 13797. 



Swertia candelabrum H, Smith in Notes Roy, Bot, Gard. Edinb., 26 : 255, pi. 34 
fig. 8 d-f (1965). 

A specimen which may belong to this species is : Bhutan, Sakden, Nyuksang La, 
3,700 m,, open hillside among rhododendron scrub, pale greenish yellow streaked 
pale reddish purple, 6 July 1934, Ludlow & Sherriff 601. It differs from the type of 
S. candelabrum, which comes from central Nepal, in the colour of its flowers, whitish 
blue in the type, and in the membraneous margin of the nectary, which is less lacini- 
ate. In other respects it is very similar ; its water-swelled ovules are ellipsoidal 
and not winged. 






HIMALAYAN SWERTIA AND VERATRILLA 



249 



Swertia crossolotna H. Smith, sp. nov. (Plate 30 ; Text-fig. 5.) 

Planta florendi tempore 22 cm. alta, S, staintonii H. Smith subsimiUs sed distat : 
caule subtereti (nee alte sulcato), nectario fimbriate (nee nudo), connectivo anther- 
arum protracto, saepe uncinatim curvato (nee subplano), floribus minoribus. 





^j^^-^-^-'v..*^ 



Fig. 5. Swertia crossolotna H. Smith : a, corolla lobe with nectary and stamen ; b, outline 
of calyx lobe ; c, margin of leaf {Ludlow 6- Sherriff 421). All x 5. 



Caulis subteres, basin versus 4-alatus, alis ciliatis ; inflorescentia pyramidalis e 
basi et nodis 3 ramosa, ramis, ramulis et pedicellis ciliato-alatis. Folia rosularia ad 
II cm. longa, lamina late lanceolata, acuta, ad 2 cm. lata, in petiolum subaequilongum 
sensim attenuata, margine crebre lacinulata, lacinulis piliformibus 0-3-0-4 mm. longis. 
Flores pallide albo-virescentes (e coUectore) ; calycis lobi subaequales, c. 7 mm. longi 
et 3-5 mm. lati, ovato-triangulares, subobtusi vel acutiusculi, margine irregulariter 
lacinulati, basin versus etiam ciliati, marginibus basalibus in alas pedicelli decurrenti- 
bus ; coroUae lobi subobovales ad 1-9 cm. longi et i-o cm. lati, superiori parte minute, 
deorsum grossius lacerulati, nectario orbiculari circum longe fimbriato ; stamina 
tubo adnata, filamentis liberis 10 mm. longis, thecis c. 3 x 2 mm. magnis, connectivo 
antherarum protracto saepe uncinatim curvato. Capsula, semina non visa. 



250 



HIMALAYAN SWERTIA AND VERATRILLA 



Bhutan : Me La, 4,300 m., open grassy hillside, pale creamy green, striped blue 
at base, 6 Aug. 1933, Ludlow & Sherriff 421 (holotype). 

Swertia splendens H. Smith, sp. nov. (Plate 31; Text-fig. 6.) 

Planta plus quam 40 cm. alta (parte basali deficiente), ex affinitate S. crossolomae 
H. Smith, a qua distat : planta multo majore, inflorescentia laxiore, floribus lateritiis, 
2*5 cm. longis (nee virescentibus, 1-7 cm. longis), coroUae lobis late obovatis sub-j 
truncatis (nee ovalibus), connective antherarum abbreviate (nee uncinatim pro- 
tracto), fimbriis nectarii inflexis i mm. longis (nee erectis 2 mm. longis). 




\\JvvJ 



^^^^\J\^Ji^^/iv3u^ou\i^^\ivi^jVl\i\JO\^ 



Fig. 6. Swertia splendens H. Smith : a, outline of calyx lobe ; b, base of corolla lobe with 
nectary ; c, anther ; d, ovary ; e, margin of leaf {Ludlow &■ Sherriff 654). All X5. 



Caulis (pars florifera solum adest) robustus, obsolete quadri-alatus, alis sparsissim« 
sed robuste ciliatis ; nodi 3 inflorescentiae pyramidalis distantes, rami infimi arcuatii 
adscendentes 2-5-flori, in nodo supremo flores plures (ad 10) cymosi. Folia rosularij 
(duo adsunt) spathulata, 7-nervia, lamina obovata, subacuta, 8 et 12 cm. longa, cj 
5-5 cm. lata, margine crebre lacinulata, lacinulis piliformibus c. 0-3 mm. longis, ii 
petiolum latum, subaequilongum attenuata, petiolo in margine sparse, basi intu^ 



HIMALAYAN SWERTIA AND VERATRILLA 251 

densius ciliato. Flores ampli lateritii 1-3-5 cm. longe pedicellati, pedicellis quadri- 
alatis glabris ; calycis tubus perbrevis, lobi triangulari-cordati, majores ad i-6 cm. 
longi, i-o cm. lati, margine minute lacinulati basin versus etiam ciliati ; coroUae 
tubus brevis, lobi late obovati, 2-3 cm. longi et 2-0 cm. lati, margine apicaliter 
minutissime lacerulati, nectario orbiculari fimbriis inflexis i mm. longis cincto ; 
stamina tube 2 mm. alte adnata, filamentis liberis 11 mm. longis, thecis 3-5 mm. 
longis, apicibus thecarum connectivo abbreviato liberis ; ovarium ovatum, stylo 
cum stigmatibus parvis 2 mm. longo ; capsula, semina non visa, ovula, ut videtur, 
subglobosa exalata. 

Tibet : Milakatong La, near Tawang, 4,000 m., open stony hillside, brick red, 
14 July 1934, Ludlow & Sherriff 6^4 (holotype). 



Sectio OPHELIA 

Swertia sect. Ophelia (D. Don ex G. Don) Gilg in Engler & Prantl, Naturl. Planzenfam. 

4 (2) : 88 (1895). 

Ophelia D. Don ex G. Don, Gen. Syst. 4 : 178 (1837). 

Swertia series Ophelia (D. Don ex G. Don) Benth. & Hook., Gen. PI. 2 : 817 (1876). 

Swertia subgen. Ophelia (D. Don ex G. Don) C. B. Clarke in Hook, f., Fl. Brit. Ind. 4 : 121 

(1883). 

Swertia franchetiana H. Smith, nom. nov. (Plate 32.) 

Swertia striata Franchet in Bull. Soc. Bot. France 46 : 322 (1900) ; non S. stricta CoUett & 
Hemsl. in Journ. Linn. Soc. LoncL, Bot. 28 : 91 (1890). 

Tibet : Hills north of Lhasa, 3,800 m., on open sandy banks, flowers pale blue 
violet striated darker, 31 Aug. 1943, Ludlow & Sherriff ggoy. Lhakang, 3,700 m., 
open hillside, very pale blue, veins mauve, 2 Sept. 1933, Ludlow & Sherriff 518. 
Kongbo, Lhatra Chu (Pangkar), 4,100 m., on dry ground, calyx green, corolla bluish 
white, filaments blue, anthers very dark brown, style green, stigma very dark brown, 
7 Sept. 1947, Ludlow, Sherriff & Elliot 15694. Kongbo, Temo La, 3,200-3,400 m., 
on grassy banks in forest, calyx green, petals pale watery blue with dark blue veins, 
two lateral lines of purple hairs at base, stamens pale blue, anthers purple, ovary 
light green at base, darker above, 7 Sept. 1938, Ludlow, Sherriff & Taylor 6gg2. 
Kongbo, Tsangpo Valley, Pe, 2900 m., in grass by streams, calyx green, corolla slaty 
blue veins darker, 23 Aug. 1947, Ludlow, Sherriff & Elliot 14436. 

Franchet described this species from western China ; it has not previously been 
recorded outside China proper. 

Swertia gracilescens H. Smith, sp. nov. (Plate 33.) 

Planta hapaxantha, S. paniculatae Wall, affinis, sed gracilior, flores minores, 
emaculati, eleganter dispersi (nee in cymas modice contracti), folia angustiora, 
rarissime ad 3 mm. lata. 



252 



HIMALAYAN SWERTIA AND VERATRILLA 



Caulis erectus ad 35 cm. altus, basin versus ad 2 mm. crassus, in lineis decurrentibus 
brevissime ciliolatus vel glabrescens, e nodis 10-13 pyramidatim ramosus, ramis 
numerosis, gracilibus, inferioribus maximis ad 10 cm. longis et 0-5 mm. crassis, 
suberectis vel subpatentibus, sursum gradatim deminuta. Folia caulina lanceolato- 
linearia, uninervia, media maxima ad 2-5 cm. longa, circ. 2 mm. lata, marginibus 
glabra. Floras albi, emaculati, 5-6 mm. longi, 5-meri, uninectariati, gracillime 4-6 
mm. longe pedicellati ; calycis tubus perbrevis, lobi attenuato-lineares, 4-5-5 mm. 
longi, basi 0-5-1 mm. lati, emucronati ; corollae tubus minus quam i mm. longus, 
lobi lanceolato-ovati, acuti, ad 5-5 mm. longi et 4 mm. lati, nectario glabro, orbicu- 
lato, margine superiori elevato ; stamina tubo 0-7 mm. alte adnata, filamentis liberis 
non dilatatis 3-5 mm. longis, thecis 0-8x0-3 mm. magnis ; ovarium anguste ovoideum 
in stylum i'5 mm. longum contractum, stigmatibus vix 0-5 mm. longis. Capsula, 
semina non visa. 




Fig. 7. Sivertia burmanica H. Smith : corolla lobe with nectary and stamen. X5. 



Nepal : Bhuji Khola, 2,100 m., open slope by track, flowers white, 15 Oct. 1954, 
Stainton, Sykes & Williams go4g (holotype). Near Gurjakhani, 2,600 m., exposed 
grass slope, flowers white, 17 Sept. 1954, Stainton, Sykes & Williams 4470. Kali 
Gandaki Valley, Taglung, south of Tukucha, 3,700 m., on grass slopes, petals and 
filaments white, anthers black, 22 Sept. 1954, Stainton, Sykes & Williams ygyo. 



I 



HIMALAYAN SWERTIA AND VERATRILLA 253 

Swertia burmanica H. Smith, sp. nov. (Plate 34 ; Text-fig. 7.) 

Swertia burmanica H. Smith ex Nilsson in Grana Palyn. 7 : no (1967), nom. nud. 

Planta hapaxantha Swertia racemosae (Griseb.) C. B. Clarke affinis et habitu similis, 
tamen distat : foliis subcoriaceis, lineari-lanceolatis, media parte (nee basin versus) 
latissimis, ad 2-5 cm. longis et 0-5-0-7 cm. latis, pagina inferiori conspicue flavescente ; 
floribus amplioribus ad 1-5 cm. longis, calyce turbinate (nee campanulato) , tubo 
calycis breviore, i •5-2-5 mm. longo (nee 2-5 mm.) ; corolla fere ad basin filamentorum 
partita, lobis late ovalibus, acutis, minute mucronulatis, margine obsolete micro- 
ciliatis, ad 13 mm. longis et 10 mm. latis, tubo 5-plo longioribus (nee sublongioribus) , 
indusio nectarii robusto, squamiforme, margine libero partim incurvante breviter 
laciniato, laciniis ecoloratis (nee tenero, magis diviso, laciniis vulgo nigrescentibus) ; 
filamentis deorsum valde incrassatis, basi approximatis sed non connatis (nee basi 
fere i mm. longe tubiforme connatis). 

Burma : Mt. Victoria Ridge, 2,600 m., a gregarious weed growing all along the 
ridge and on turf slopes in thousands, flowers 5-petalled, purple, 26 Oct. 1956, 
Kingdon-Ward 2 2^47 (holotype). 

On the type sheet there are 3 individuals present. Two of them are more or less 
densely ciliate on the wings of the stem, on the surface and base of the under-side of 
the leaves, and on the calyx. The third specimen is nearly glabrous except on the 
wings of the stem. The degree of hairiness seems to be of small taxonomic import- 
ance, as is also the case in 5. racemosa (below). 

Swertia racemosa (Griseb.) C. B. Clarke in Hook, f., Fl. Brit. Ind. 4 : 124 (1883). 

Swertia racemosa Wall., Numer. List : 154, n. 4377 (1831), nom. nud. 
Ophelia racemosa Griseb., Gen. & Sp. Gentian. : 319 (1838). 

Kingdon-Wardia codonopsidoides Marquand in Journ. Linn. Soc. Lond., Bot. 48 : 207 
(1929) ; in Hook. Icon. PI. 32 : t. 3163 (1932). 

Tibet : Kongbo, Pa La (Tra La), 3,700 m., in colonies in damp shady situations, 
flowers pale dull purple, 26 Sept. 1924, Kingdon-Ward 6205 [isotype of Kingdon-Wardia 
codonopsidoides). Kongbo, Bo La, 4,300 m., under bushes in light forest, calyx 
light green, corolla purplish white, filaments pale blue, anthers black, 9 Sept. 1947, 
Ludlow, Sherriff & Elliot 15705. Kongbo, Nambu La, 4,400 m., on grassy banks 
in juniper forest, corolla pinkish mauve, 31 Aug. 1938, Ludlow, Sherriff & Taylor 
6g5i. Tsari valley, 4,000-4,300 m., on alpine meadow slopes, 11 Sept. 1935, King- 
don-Ward 12324. Tsari, Podzo Sumdo, 3,500 m., by the side of a stream in dense 
forest, corolla palest mauve, 5 Sept. 1936, Ludlow & Sherriff 2544. 

Nepal : Deoli Patan, S.E. of Jumla, 3,800 m., growing on open slopes, corolla 
mauve, stamens blue, 5 Sept. 1952, Polunin, Sykes & Williams JiSg. Rohagaon, 
Suli Gad, 3,400 m., open slopes, corolla mauve, nectaries at base of lobes green with 
fringe of dark blue hairs, filaments blue at apex, anthers brown, style blue, 14 Sept. 
1952, Polunin, Sykes & Williams 3386. Near Seng Khola, 4,100 m., grass gullies, 
flowers pale mauve, 3 Oct. 1954, Stainton, Sykes & Williams 4688. Near Phagune 



il 



254 HIMALAYAN SWERTIA AND VERATRILLA 

Dhuri, 3,500 m., sheltered gully, flowers lilac, 12 Oct. 1954, Stainton, Sykes & 
Williams 4799. Above Dogadi Khola, 4,600 m., exposed earth pockets near ridge, 
flowers mauve, calyx and pedicels dark purple, 12 Aug. 1954, Stainton, Sykes & 
Williams 3834. Same loc, 4,000 m., among dwarf Rhododendron, flowers light 
purple, I Oct. 1954, Stainton, Sykes & Williams 4661. Dhorpatan, 2,900 m., among 
shrubs on bank, flowers mauve, 16 Oct. 1954, Stainton, Sykes & Williams 9056. 
Above Sauwala Khola, 3,400 m., grass slopes on ridge, flowers mauve, 13 Sept. 1954, 
Stainton, Sykes & Williams 4381. Same loc, 4,000 m., among short grass on ridge 
top, abundant, flowers mauve, 13 Sept. 1954, Stainton, Sykes & Williams 4389. 
N.W. of Gurjakhani, 3,500 m., grassy slope, flowers white, 20 Sept. 1954, Stainton, 
Sykes & Williams 4502. Same loc, 3,700 m., near grazing encampment, abundant, 
flowers dark mauve, 20 Sept. 1954, Stainton, Sykes & Williams 4512. Kali Gandaki 
Valley, Samargaon, north of Tukucha, 4,900 m., on steep grass slopes, calyx mauvish 
green and hairy, corolla, filaments and anthers pale mauve, 16 Aug. 1954, Stainton, 
Sykes & Williams 7249. Kali Gandaki Valley, Sangdah, north of Tukucha, 4,100 m., 
open grass slopes, calyx green, corolla, filaments and anthers whitish blue, 18 Aug. 
1954, Stainton, Sykes & Williams 7309. Kali Gandaki Valley, Tukucha, 3,700 m., 
open grass slopes, calyx greenish mauve, corolla and filaments pale mauve, lower 
leaves reddish green, 9 Sept. 1954, Stainton, Sykes & Williams 7735. Same loc 
3,700 m., open grass slopes, calyx greenish white, corolla and filaments white, 
anthers black, 9 Sept. 1954, Stainton, Sykes & Williams 7737. Same loc, 3,100 m 
open grass slopes, corolla white, 15 Oct. 1954, Stainton, Sykes & Williams 8157. 
Kali Gandaki Valley, Taglung, south of Tukucha, 3,700 m., open grass slopes, petals 
and filaments white, anthers black, 20 Sept. 1954, Stainton, Sykes & Williams 7945 
Sabze Khola, 4,000 m., among low bushes in moist ground and by stream side,, 
flowers pale lilac or mauve, leaves and stem stained red, 6 Sept. 1950, Lowndes L1491. 
Khangsar, 4,300-4,600 m., open dryish hillside, in great quantity, flowers pale lilac 
or mauve, leaves and stem stained red, 9 Sept. 1950, Lowndes L1502. Bimtakothi, 
4,000 m., damp hillsides, flowers whitish, 12 Sept. 1950, Lowndes L1539. Baraha 
Pokhri ridge, 4,000 m., open grassy hillside, flowers greyish white, 19 Sept. 1950 
Lowndes L1526. Mardi Khola, Annapurna Himal, 3,500 m., among shrubs, corolla 
pale blue, 17 Sept. 1954, Stainton, Sykes & Williams 8465. Rambrong, Lamjung 
Himal, 3,700 m., on open hillside, corolla pale blue, 18 Sept. 1954, Stainton, Sykes & 
Williams 8631. Kyangjin Ghyang, 4,000 m., grassy ground, pale purple, 0. Polunin 
1909. " Gossain Thar " [? = Gosainkund], Wallich 4377 (isotype of Swerti 
racemosa). Gosainkund, 4,900 m., 28 Oct. 1935, F. M. Bailey's collectors 60. E. of 
Num, Maghang Khola, Arun valley, 3,500 m., on open slopes, corolla mauve, 15 Sept. 
1956, Stainton 1737. Sabel, 4,000 m., 18 Sept. 1937, K. N. Sharma 63/94. Chocho, 
3,500 m., 27 Sept. 1937, K. N. Sharma 56/94. Gupchet, 3,700 m., 13 Sept. I935r 
F. M. Bailey's collectors s.n. Chilung Pati, 4,100 m., 25 Oct. 1935, F. M. Bailey's, 
collectors 49. Rolwaling valley, 3,800 m., 28 Sept. i960, Swan R 90. -J^ 

SiKKiM : Changu, 3,700 m., on loamy and sandy banks, blue, 25 cm. tall, 27 Sept. 
1 913, Cooper 934. 

Bhutan : Lingshi, 4,100 m., shrubbery, flowers dull purple, 24 Sept. 1967, J^ow^sj 



HIMALAYAN SWERTIA AND VERATRILLA 255 

Lyon 5135. Thimpu, 3,100 m., on dry loam, blue to purple flowers, 25 cm. tall, 
slender, 5 Nov. 1914, R. E. Cooper 3349. Kopub, 3,400 m., in mould and leaves, 
etc., at edge of forest, flowers pale blue and white, 24 Sept. 1914, R. E. Cooper 
2141. Lhabja, 3,800 m., on moist soil under forest, flowers blue, 30 cm. tall, 22 
Sept. 1914, R. E. Cooper 2i8y. Waitang, Tsampa, 4,000 m., in dense vegetation 
of steep hillside, flowers pale pink with a tinge of blue, up to 40 cm. tall, 19 Sept. 
1949, Ludlow, Sherriff & Hicks 19753. Tolegang, Tsampa, 4,300 m., among dwarf 
rhododendron on the open hillside, corolla pale slaty blue, 10 Sept. 1949, Ludlow, 
Sherriff & Hicks 19730. 

Assam : Near Orka La, 3,700-4,000 m., alpine turf slopes along the ridge, flowers 
purple, 26 Sept. 1938, Kingdon-Ward 14292. 

The genus Kingdon-W ardia is said by its author to be intermediate in position 
between Gentiana and Swertia, differing from the former by the presence of a fimbriate 
nectary at the base of each corolla lobe and from the latter by " the conspicuous 
calyx tube surmounted by triangular lobes and the well-marked corolla tube." The 
distinction from Gentiana certainly holds true, but the separation from Swertia is 
not convincing. Marquand was obviously not acquainted with Swertia racemosa, a 
species little known in 1928 and represented in the Kew herbarium at that date by 
only the Wallich type specimen and by the specimen from Tibet to which Burkill 
had given the manuscript name Swertia codonopsidoides. 

The only statement in the diagnosis of Kingdon-W ardia that does not fit Swertia 
racemosa is : " corolla circa ad ^ lobata ". This statement is, however, not correct ; 
a more exact one would be : corolla ad ^ vel pauUo ultra lobata. The figure subse- 
quently published by Marquand in Hooker's Icones shows the lobes and the tube to 
be of equal length, but the previous incorrect statement is repeated in the text. In 
the isotypes that I have seen my own measurements show that the lobes are in fact 
constantly a trifle longer than the tube, exactly as in Swertia racemosa, with which 
species Kingdon-W ardia codonopsidoides is undoubtedly conspecific. 

Swertia racemosa has a slightly isolated position in the genus, differing from other 
species of Swertia sect. Ophelia by the less divided corolla and by the well developed 
calyx tube, but it is not divergent in any other respect. Its affinity with such species 
as S. ciliata (G. Don) Burtt and S. hispidicalyx Burk. is indicated by the shape of the 
nectary, stamens and style. It is even closer to 5. hurmanica, which is very similar 
in appearance. The calyx and deeply cleft corolla of that species conform to the 
normal pattern found in S. sect. Ophelia and it thus links 5. racemosa with the other 
members of the section. 

Plentiful material of Swertia racemosa has been collected in recent years, especially 
by the British Museum expeditions to the Himalayas. In that herbarium there are 
32 gatherings from Nepal, i from Sikkim, 6 from Bhutan, 5 from S.E. Tibet and i 
from Assam. This ample material shows that S. racemosa has an unusually wide 
range of variability. The corolla lobes vary from distinctly to only slightly longer 
than the tube ; the calyx varies from 5 to 2 mm. in length, the calyx lobes are usually 
much longer than the tube and distinctly unequal, but they may also be shorter than 
the tube and of subequal size, and the angles between the lobes vary from broadly 



256 HIMALAYAN SWERTIA AND VERATRILLA 

rounded to narrowly acute ; the stigma is sometimes subsessile, but commonly the 
style is more or less prolonged and may reach a length of 5 mm. ; the petals average 
in length 11 mm. (5-14) ; the calyx, the margins and upper surface of the leaves vary 
from densely or sparingly long- or short-ciliate to perfectly glabrous. All these 
variations seem to be independent of each other, and the extremes are connected by 
intermediate stages. To name such indistinct forms would be meaningless. It may 
nevertheless be mentioned that really long-styled specimens are not found outside 
Nepal, but this may be due to the fact that the plant has been collected over 30 times 
there and only 13 times elsewhere. 

VERATRILLA (BaiUon) Franchet 
Veratrilla (BaiUon) Franchet in Bull. Soc. Bot. France 46 : 310 (1900). 
Swertia sect. Veratrilla Baillon in Bull. Mens. Soc. Linn. Paris i : 730 (1888). 

At a session of the Linnean Society of Paris on 6 April 1888, Baillon demonstrated 
a gentianaceous plant collected by Delavay in Yunnan. He gave a description, 
pointing out that it had certain characters in common with Jceschkea, Obolaria, 
Bartonia and Gentiana, but concluded "... Swertia proprement dits, et c'est a ce 
genre que le Veratrilla, si singulier qu'il puisse paraitre, devra se rapporter comme 
section." This must be regarded as valid publication of the name Swertia sect. 
Veratrilla. The sectional epithet was chosen on account of the resemblance of this 
plant to Veratrum L. Baillon proposed no specific name for this plant. 

In the following year, Baillon (Hist. PI. 10 : 142 (1889)) expressed uncertainty 
about the taxonomic status of the Delavay plant, saying in a footnote to Swertia 
" aut generis sectio, aut generis proprium erit Veratrilla H. Bn. . . ., cujus fiores 
dioeci, saepe 4-meri . . . Species asiaticae multae novae banc plantam cum Frasera 
et Swertiis legitimis nihilhominus connectunt." 

Franchet, who had at an earlier date, probably in 1885, given a manuscript name 
under Swertia to the Delavay specimen in the Paris herbarium, obviously became 
convinced that it represented a separate genus and published it in 1900 at that rank 
and, ignoring his earlier manuscript epithet, published the name Veratrilla haillonii 
for the species. I quote his diagnosis in extenso ; there are two statements in it 
that should be altered. It seems as if Franchet had by that time lost interest in the 
plant ; the description is not of the same high class as one is used to from him : 

" Veratrilla, an generis sectio an genus proprium Baillon, Hist, des pi. X, 142 et 
Bull. Soc. Linn, de Paris, 729. 

" Fiores diceci, saepius 4-meri ; corolla tubuloso-rotata, staminibus inferne cum 
tubo connatis juxta sinum tantum liberis, brevibus ; stylus brevis (vix i mm.) 
stigmate bilobo, lobis ovatis patentibus ; capsula ovata, breviter attenuata ; semina 
late ovata, levia. — Planta glabra, elata veratriformis ; folia inferiora longe petiolata, 
omnia opposita, anceolata [sic], inferiora vaginata ; infiorescentia e racemis densis 
pyramidam longam, angustam efformantibus ; corolla luteoviridis, 6 mm. longa et 
lata. 

" V, Bailloni sp. nov. 



» 



HIMALAYAN SWERTIA AND VERATRILLA 257 

Hab. — La Chine occidentale, Yunnan, dans les broussailles, au-dessus du col de 
Yen-tze-hai, alt. 3500 metres (Delavay) ; prairies humides de Tsang-chan, pres de 
I'ar^te (Delavay n. 147). 

" Genre remarquable, a cause de sa dioicite ; les etamines sont ins^rees comme celle 
des J cBSchkea." 

Baillon said in his original account " ses fleurs sont ordinairement tetrameres, plus 
rarement pentameres, principalement dans le pied femelle " and in the Histoire des 
Plantes " flores . . . saepe 4-meri ", and Franchet said " flores . . . saepius 4-meri ", 
but I have examined 16 different collections and have never found a single flower 
that was not tetramerous. Kingdon-Ward, however, recorded pentamerous flowers 
in a field note (see below). 

Franchet 's description of the seeds reads "semina late ovata, levia". A more 
accurate one would be : semina complanata, ambitu subovalia, circum anguste alata 
c. 3x 1-5-2 mm. magna, ala 0-3-0-5 mm. lata. 

The great difference in appearance between the male and female plants must also 
be pointed out. The female plant has the long narrow pyramid of Franchet's 
description ; each bract supports one or two short upright branches bearing fewer 
than ten pedicelled flowers. The inflorescence of the male plant is robust and 
spreading, several hundreds of small flowers without pedicels being massed together 
on the stout and more or less patent branches. This difference is well demonstrated 
by the specimens of this species from China that are illustrated as plates 35-36. 
Those from Tibet or Assam are not so satisfactory for photography. 

Veratrilla baillonii Franchet in Bull. Soc. Bot. France 46 : 311 (1900) (" bailloni "). 
(Plates 35, 36.) 

Swertia mekongensis Balf. f. & Forr. in Notes R. Bot. Gard. Edinb. 4 : 80, t. 19 (1907). 

S.E. Tibet : Mt. Kenichunpo, Salween and Irrawaddy divide, 4,600 m., limestone 
crags, alpine region, plant fleshy, flowers yellow, May-July 1932, Rock 2ig38. 

Assam : Orka La, Bhutan frontier, 4,000 m., pastures, flowers 4-merous, sometimes 
5-merous, Kingdon-Ward 1372$. 

Veratrilla baillonii had not previously been recorded outside China proper, the 
identity of Swertia mekongensis with it not having been recognized. It is now 
recorded for S.E. Tibet and Assam. 

Veratrilla burkilliana (W. W. Smith) H. Smith comb. nov. (Plate 37.) 

Swertia burkilliana W. W. Smith in Journ. Asiat. Soc. Bengal, new ser. 7 : 78, t. i, fig. i-iii 

(1911). 
Veratrilla burkilliana H. Smith apud Nilsson in Grana Palyn. 7 : 113 (1967), nom. nud. 

S.E. Tibet : Tsari, Tomtsang, 4,300 m., on open grassy hillside, corolla dull 
purple margined with green and slightly streaked with greenish purple, anthers blue 
purple, 23 June 1936, Ludlow & Sherriff 2204. Same loc, 4,300 m., on open grassy 
hillside, flower over, 13 Sept. 1936, Ludlow & Sherriff 2578. Chubumbu La, Langong, 



258 



HIMALAYAN SWERTIA AND VERATRILLA 



4,000 m., common in open swampy meadows, corolla pale green, tinged dull slate 
blue at base, 2 June 1938, Ludlow, Sherriff & Taylor 3943. 

SiKKiM : Changu, 3,700 m., in turf on lake side, green flowers, black marks, blue! 
stamens, i July 19 13, R- E. Cooper g2. Same loc, 3,700 m., on lake border, 
greenish flowers with spots, 20 Aug. 1913, R. E. Cooper 565. 

Bhutan : Pung La, 3,700 m., open marshy places near top of pass, flowers blue] 
marked with green, 9 July 1949, Ludlow, Sherriff & Hicks 20903. 

This second species of Veratrilla agrees exactly in all generic characteristics with] 
V. baillonii, but it is quite distinct specifically. The plant is much smaller and less] 
robust, only 15-30 cm. tall as against 30-50 cm. The cauline leaves are shorter and! 
relatively broader. The seeds are slightly smaller, nearly circular in outline and! 
with a slightly broader wing. The difference in appearance between male and female 
plants is strongly marked in this species also. The excellent figure accompanying! 
the original description of the species represents the female plant. In the male the] 
inflorescence is spreading, up to 4 cm. wide by 6 cm. long. 



Dr. H. Smith 

Institute of Systematic Botany of the University of Uppsala 

Botanical Garden and Botanical Museum 

Uppsala i 

Sweden 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 22 




Coll. J. D. A. ST'AINTON 



N.,.27Gb. 



Swertia kiharae Kitam. {Stainton 2yo8 in Herb. Brit. Mus.) 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 23 




Swertia hooker i C.B. Clarke 



Sivertia hookeri C. B. Clarke [Ludloiv, Sherriff &■ Elliot 14376 in Herb. Brit. INIus. 



< 



r 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 24 




Swertia acaulis H. Smith {Lall Dhowj 02gi ; holotype in Herb. Brit. Mus. 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 25 




n.oRA or 6RyT<^iH 






Alt, j3 coe' 



^fCewiM, ■5K.M ^«^ ^^ ^ "P^ 



^^u^Ul**^ 



Coll, r-'. LtoU>» 



Sivertia grandiflova H. Smith {Ludloiv, Sherriff 6- Hicks 2oy32 ; holotype in Herb. Brit. 

Mus.) 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 26 



hnrnw MtMoM (»»«»«»» >i»M»»| 
IiiiIiw'""I""I'«iI«mHi«i1h«IiiJ 




IL„ 




FtOK* OF 6l*UT^l1 






' 19323 






Colli F H.BI..W, C.Shi ii«iri* J II. llir.M 



Sivertia virescens H. Smith [Ludlow, Sherriff &- Hicks 19323 ; holotype in Herb. Brit. Mus.) 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 27 



f 1.0MA Of eHUTno . \. 19414 



Ml 1<IA:k d,m, 



I 




Swertia pseudo-hookeri H. Smith [Ludlow, Sherriff cS- i/ic^5 jp-^i"^ ; holotype in Herb. Brit. 

Mus.) 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 28 




FUJRA OF EASTERN NEPAL 






Alt , nut. 



0)U. J. D. A. STAiNTON 



No. 755 



Swertia staintonii H. Smith [Stainton 753 ; holotype in Herb. Brit. Mus.) 



< 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 29 



1 1 1 n 1 1 1 1 1 1 1 ! 1 1 11 1 1 ! ] If 
' 11 ' 2 

BUTtSH MusaiM (Natvuo. Histoky) 
Photograw* 




A' 



Britl'.h MusiMim LMw.lill>'» 

FLORA Oh ASSAM 1S..> 
(;«pl, F. Kingiluii-W.rd 



Swertia assamensis H. Smith [Kingdon-Ward 14008 ; holot^^pe in Herb. Brit. Mus. The 
printed number on the ticket has been deleted in bhie ink and 14008 substituted, but 
this has not reproduced on the photograph.) 



l. 



i 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 30 




"^ <^t,.**#^r»^* -<'4'» 



,^J 



HOLOTY PE 
Swertia crossoluma H. Smith 



Ottt*X4l PUj^ f'frrA./u P , 



Swertia crossoloma H. Smith {Ludlow cS- Sherriff 421 ; holotype in Herb. Brit. Mus. 



V 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 31 



I 




Swertia splendens H. Smith {Ludlow 6- Sherriff 654 ; holotype in Herb. Brit. Mus. 



V 



I 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 32 




f. LUDLOW-G 5HERKIFF. 

s.r'&i- 






Swertia franchetiana H. Smith {Ludlow &• Sherriff 518 in Herb. Brit. Mus. 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 33 



' ' ' 1 1 ' I " I " I " I " I " I " 

•nnn MunuiK (N*niUL HntoaTl 
PHorofaiArH 




HOLOTYPE 
Swertia graeU««cen» 
~" H". Smith 



FLORA OF We^ Ni,,eiol,.<^. 

■A WIJa. Alt. l|0*o|j^- !>«» l5j iS)| (<j$L. 



*A^W . SJatA <) iJ^ixivwA . 



Swertia gracilescens H. Smith {Stainton, Sykes &> Williams go4g ; holotype in Herb. Brit. 

Mus.) 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 34 



ti?'r/ 




V. KINCDON WARD 



22747 



Swertia bimnanica H. Smith [Kingdon-W ard 22^4^ ; holotype in Herb. Brit. Mus.) 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



I 







i,-,f f /^/O 

^/t/*a^i£a '-'^Kx^iJ^tn^ ..^Cv*,^^ _ 



t(gft i4j/eKt^ ^njf tfM^f ^^-i f-^'yV 



Veratrilla baillonii Franchet, male plant {Forrest 6983 in Herb. Brit. Mus. 



Bull. Br. Miis. nat. Hist. (Bot.) 4, 6 



PLATE 36 




I 



■ilah gr&«n. 



iu\f oer.i iitj.ae . 



'".\jiA.jiCii*. 0»ia<„-. "'k^-K^^ 



Ciliocnii T. T. TC 



Veratrilla baillonii Franchet, female plant {Yii 13514 in Herb. Brit. Mus. 



J 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 6 



PLATE 37 




KuiR.^ or |{^i,*"^M 



2090.3 



U lA (e $H01f>^ ,v,, j/^,„, „.,, ^Lj^ 






^ii./-% ;^ ;, V A //< , 



Veratrilla hurkilliana (W. W. Smith) H. Smith, male plants {Ludlow, Sherriff &• Hicks sogoj 

in Herb. Brit. Mus.) 



\ 



.';-%•'' »T»^/-':/>1^Stfle/S^*^ 



Printed b Great Britain by 

Alden & Mowbray Ltd 
at the Alden Press, Oxford 



A SURVEY OF THE TROPICAL 

GENERA OPLONIA AND 
PSILANTHELE (ACANTHACEAE) 



W. T. STEARN 



BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 7 

LONDON : 1971 



A SURVEY OF THE TROPICAL GENERA 
OPLONIA AND PSILANTHELE (ACANTHACEAE) 




BY 

WILLIAM THOMAS STEARN 



)o^ 



pp. 259-323 ; 18 Text-figures ; Plates 38-47 



BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

BOTANY Vol. 4 No. 7 

LONDON : 1971 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 
within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 7 of the Botanical 
series. 



World List abbreviation 
Bull. Br. Mus. not. Hist. (Bot.). 



/' >* 



Trustees of the British Museum (Natural History), 197 1 



trustees of 
the british museum (natural history) 



Issued 10 August 1971 



Price £2.60 



A SURVEY OF THE TROPICAL GENERA 
OPLONIA AND PSILANTHELE (ACANTHACEAE) 

By WILLIAM T. STEARN 

SUMMARY 

On grounds of priority the generic name Oplonia Raf. (1838) replaces Anthacanthus Nees 
(1847) ; both have the same type, Justicia spinosa Jacq. The genus Psilanthele Lindau (1897) 
is restricted to its type-species, P. eggersii Lindau of Ecuador ; the Peruvian and Jamaican 
species put in Psilanthele by Lindau are transferred to Oplonia, the supposed distinction between 
the " Rahmenpollen " of the one and " Spangenpollen " of the other being illusory. Forsythi- 
opsis Baker (1883), hitherto considered endemic to Madagascar, is here united with Oplonia, as 
they agree in habit, floral structure, capsule, and pollen. The genus Oplonia thus extended 
consists of one species in Peru, eight in the West Indies, none in Central America, and five in 
Madagascar, which, like Cuba and Jamaica, is a major centre of endemism. Recognition of taxa 
within the group is very difficult owing to extensive but seemingly uncorrelated variation in 
habit of growth, pubescence of shoots, and size and shape of leaves. The spines provide useful 
diagnostic characters. Floral dimorphism, i.e. the occurrence of long-stamened short-styled 
flowers and of short-stamened long-styled flowers on different individuals of the same taxon in the 
same district, is demonstrated, possibly for the first time in the Acanthaceae ; otherwise the 
genus manifests very little floral diversity. Most species grow in dry open habitats as part of 
scrub forest. Their evolution would appear to be from a mesophytic spineless South American 
ancestor, resembling the small-flowered species of Odontonema, by progressive reduction in size 
of leaf, by contraction of the inflorescence from a narrow panicle or raceme into a few-flowered 
axillary fascicle or a solitary axillary flower, and by conversion of lateral branches into spines, 
by which changes they became xerophytic. The most primitive species on this assumption are 
0. grandiflora of Peru and O. jamaicensis of Jamaica. Parallels in distribution with Neriacanthus, 
Heppiella, Huerta and Purdiaea indicate ancient floristic links between the Greater Antilles and 
north-western South America independent of Central America and the Lesser Antilles and 
suggest that a major part of the Greater Antillean flora may be descended from plants of a land- 
mass now sunk which occupied the position of the present Caribbean Sea. 

The presence of congeneric species in tropical America and Madagascar but not in continental 
Africa, as exemplified by Oplonia, Oliganthes, and Ravenala, is attributed to continental drift 
and subsequent extinction on the African mainland, as there exist genera with ranges inter- 
mediate between the extreme discontinuity of these and the continuous range of, e.g., Sabicea. 

New nomenclatural combinations are: Oplonia acicularis [Justicia acicularis Swartz), O. 
armata [J. armata Swartz), O. grandiflora (Psilanthele grandiflora Lindau), O. linifolia {Forsythi- 
opsis linifolia Benoist), O. microphylla {J. microphylla Lam.), O. minor [F. minor Benoist), 
O. minor var. meridionalis {F. vincoides var. meridionalis Benoist), O. minor var. vestita {F. vin- 
coides var. vestita Benoist), O. nannophylla {Anthacanthus nannophyllus Urban), O. purpurascens 
{A . purpurascens Griseb.), O. vincoides (/. vincoides Lam.). O. tetrasticha var. polyece from Oriente 
province, Cuba, and O. acuminata and O. puberula, both from Madagascar, are described as 
new ; Psilanthele minor Lindau is renamed O. armata var. pallidior. Descriptions, synonymy 
and distributional data are provided. 

As a guide to the capacity of computer-aided taxonometric (numerical taxonomic) methods to 
facilitate the sorting of relatively intractable specimens into groups definable and acceptable as 
taxonomic units, two of these methods (using programmes CLASP and Taximetric Similarity 



262 



A SURVEY OF THE TROPICAL GENERA 



Graph-clustering) were applied to Oplonia specimens providing a matrix of 24 unweighted 
characters. The results, expressed as stages of clustering, as a dendrogram and as a minimum 
scanning tree converted into a pictorialised scatter diagram, demonstrated their value as aids to 
understanding assemblages wherein most distinctions lie in a diversity of minute characters 
neither simultaneously observable nor easily correlated. The stage-by-stage clustering, with 
progressive reduction of the degree of similarity necessary for admission of specimens to clusters, 
portrayed graphically by the Taximetric Similarity Graph-clustering programme, commended 
itself as a time-saving substitute for part of the process of searching for correlations. 



HISTORICAL INTRODUCTION 

The genus Oplonia is a small group of woody much-branched usually spiny small- 
flowered tropical plants belonging to the family Acanthaceae subfamily Ruelloideae 
tribe Odontonemateae. The flowers are solitary in the leaf-axils or borne in com- 
paratively few-flowered axillary fascicles with minute bracts, this much-reduced 
inflorescence distinguishing Oplonia from Odontonema Nees {Thyrsacanthus Nees), 
which has terminal racemes or panicles. The calyx, as in Odontonema, is cleft into five 
narrow segments. The corolla (Text-fig. i, B) is zygomorphic, usually lilac or red, 




Fig. I. Oplonia armata var. pallidior Stearn : A, Habit, x 2/3 ; B, corolla, x i ; C, 
flower of long-styled form cut lengthwise, x 2 ; D, pollen grain, x 300 ; E, capsule, X i 
{Harris 10257). 

but sometimes white, with a slender tube and a two-lipped limb, the upper lip two- 
lobed, the lower lip three-lobed. There are two stamens, which may be exserted or 
included, and two staminodes. One species, Oplonia grandiflora, grows in northern 



OPLONIA AND PSILANTHELE (ACANTH ACEAE) 263 

Peru ; the others are West Indian or Madagascan. Most of these can be distin- 
guished from other West Indian members of the Acanthaceae by their opposite 
axillary spines (Text-fig. 2), which are modified shoots, not modified leaves as are the 
paired spines of Barleria and Barleriola. They supply the major taxonomic features 
of the species and vary from the forbidding rigid four-ranked stiletto-like thorns up 
to 3 cm. long of 0. tetrasticha to the short prickles, sometimes under 4 mm. long, of 0. 
nann&phylla. The generic names Opionia and Anthacanthus and the epithets acicu- 
laris, armata, polyece and spinosa all refer to this armature. 

Although pre-Linnaean collectors gathered a few specimens of Opionia microphylla 
in Antigua and Jamaica, the first species to be scientifically recorded was 0. spinosa 
collected by Nicolaus Joseph von Jacquin near Port-au-Prince, Haiti, in 1757-58 and 
published by him in 1760 3.s Justicia spinosa (PI. 43 a). On this species Rafinesque 
founded the genus Opionia in 1838, his characteristically contracted diagnosis being 
as follows : 

" 987. Oplonia R. (weapons) diff. Justicia, cor. hypocraterif, tubo tereto, 
limbus 5 lobo, bilabiato, lab. sup. bilobo piano. Type Opl. spinosa R. Justicia 
do. L. auct. Prickly shrub, and probably all the prickly American sp. J. micro- 
phylla, armata, acicularis &c. " 
Probably Rafinesque never saw any of the plants concerned, which are all West 
Indian and were then rare in herbaria. However he never hesitated about founding 
new genera on the illustrations and descriptions of more cautious authors* and he 
could have drawn more than enough information for his diagnosis from Vahl's 
Enumeratio Plantarum 1 : 167-169 (1804), where /. spinosa, J. microphylla, J. armata 
and J. acicularis are described. The epithets spinosa, armata and acicularis pre- 
sumably suggested the generic name Opionia, derived, like the name Oplopanax 
(Torr. & Gray) Miq., from the Greek ottXov, " implement of war, weapon, armour ", 
and thus cognate with the word " hoplite " ; H opionia would indeed have been a 
better rendering. Rafinesque's Flora Telluriana (Philadelphia, 1837-38) is a very 
rare work, of which apparently only two copies were available in Europe until the 
publication in 1946 of a photographic facsimile. Certainly C. G. Nees von Esenbeck 
knew nothing of it when in 1847 he separated these same prickly species horn. Justicia 
and Eranthemum and named the genus Anthacanthus, from dvdos " flower " and 
oLKavdos, " spine ". Nees distinguished within Anthacanthus a section Armati 
consisting of the spine-bearing West Indian species {A . spinosus, A . armatus, A . 
acicularis, A. microphyllus, A. lycioides, A. cuneatus, A. emarginatus and a " species 
dubia ", A. sprengelii, actually a member of the Rubiaceae) and a section Inermes 
consisting of unarmed species {A. vincoides from Madagascar, A. sinuatus and A. 
repandus from the New Hebrides) which could well have been left in Eranthemum 

* In his Flora Telluriana 4 ; 64, n.986 (1838), Rafinesque with customary recklessness also published 
a genus Crateola to include two little-known species, " Types the J. vincoides, J. parviflora Ortega non 
Retz.", of which he had read the brief descriptions in Vahl, Enum. Plant, i : 166-167 (1804), and charac- 
terized it as "difl. Justica, cor. hypocraterif. tubo brevis basi inflato, limbo piano 5 lobo vel 5 partito 
equalis ". As Vahl described the corolla of Justicia vincoides Lam. as " infundibiliformis "(its tube is 
in fact narrowly cylindric and not swollen at base) but stated that /. parviflora Ortega had "corollae 
tubus basi ampliatus ", I accordingly designate /. parviflora as the lectotype of Crateola Raf. Nees- 
placed this obscure species in the genus Dipteracanthus Nees (1832), although Hemsley suggested that 
it might belong to Hemigraphis Nees (1847). The name Crateola thus typified cannot displace either 
Opionia or Forsythiopsis ; should it compete with Hemigraphis, the latter name should be conserved. 



264 A SURVEY OF THE TROPICAL GENERA 

(see p. 320) as accepted then.* In 1920 Britton and Millspaugh designated Antha- 
canthus spinosus [Justicia spinosa Jacq.) as the lectotype of Anthacanthus. Since, 
however, the genus Anthacanthus as originally circumscribed by Nees included the 
designated type of Oplonia Raf., the name Anthacanthus was technically a super- 
fluous name when published and its type has always been Justicia spinosa Jacq. 
(see Int. Code Bot. Nomencl. Art 7, Note 4.) 

Subsequent botanists dealing with the family Acanthaceae have maintained the 
genus and, with the exception of Otto Kuntze, have adopted the name Anthacanthus 
for it. In 1839 Endlicher divided the genus Eranthemum into three groups of un- 
specified rank, namely Euranthemum of the East Indies, Planetanthemum of Africa 
and Australia and Hesp eranthemum of the West Indies. He did not list their species 
but his group Hesp eranthemum, comprising " frutices antillani spinosi " with " pedun- 
culi axillares, subuniflori ", obviously refers to the group typified by Eranthemum 
spinosum (Jacq.) R.Br, ex Roem. & Schult., i.e. Justicia spinosa Jacq. Kuntze 
raised this group to generic rank in 1891 as Hesperanthemum, thereby creating a 
superfluous new name. At the same time, obviously knowing nothing about the 
plants themselves covered by this nomenclatural exercise, Kuntze described as 
Jasminum coeruleum the only one he himself had collected on St. Thomas in the West 
Indies. As pointed out by Green (1969), there are no Jasminum species native to 
the New World, J.fiuminense VeU. being a naturalized African species, and Kuntze's 
Jasminum coeruleum is here regarded as conspecific with Opionia spinosa. 

Although the genus Oplonia is easily recognised and defined, the species are far 
otherwise. Gustav Lindau in 1895 described it as having " 6 (oder mehr) schwer zu 
trennende, westindische Arten ". The difficulty of separating the nine species 
named by earlier authors led him in 1900 to reduce them to three distinguished on 
characters so unreliable that he might just as well have been content with one. 
Lindau regarded A . armatus as a species endemic to Jamaica, which it is, but he 
treated A . acicularis, in fact also a Jamaican endemic, as " eine sehr variabele Species " 
extending from the Bahamas over Cuba, Jamaica, Hispaniola, St. Thomas and 
Antigua to Guadeloupe and Marie Galante. To A. spinosus he attributed a similar 
range, from the Bahamas over the Greater and Lesser Antilles to Guadeloupe and 
Martinique. Specimens with puberulous flowers he referred to the one, those with 
glabrous flowers to the other, a procedure of no value whatever. 

Meanwhile Lindau described in 1897 a new genus Psilanthele, based on a new 
species , P. eggersii, collected by Baron Henrik Eggers {Eggers 15129) in the province 
Manabi of Ecuador, which has very small flowers, the corolla being about 4 mm. long, 
in io-20-flowered raceme-like panicles terminating axillary shoots (PI. 47). The 
generic name, unexplained by Lindau, is evidently from ifjtXos, " smooth, bare ", 
dvdi^Xr], " blossom, panicle ". Since he considered its pollen to be of the " Rah- 
menpoUen " type typical of the Graptophylleae, whereas Anthacanthus and Odonto- 

* As accepted by Nees, the genus Eranthemum L., of which the type species is E. capense L., included 
not only species of Eranthemum sensu stricto [Daedalacanthus T. Anderson) but also of Pseuderanthemum 
Radlkofer (" Eranthemum " sensu T. Andersoni, non L.), as explained by Stapf in Bot. Mag. 135 : t.8239 
(1909) and Ruspolia Lindau (cf. Milne-Redhead in Kew Bull. 1936 : 255-274 (1936) ). Oplonia, Psilanthele 
and Odontonema are related to Pseuderanthemum and not to Eranthemum, of which one species, E. 
pulchellum Andrews (E. nervosum (Vahl) R. Br. ex Roem. & Schult., Daedalacanthus nervosus (Vahl) 
T. Anderson), is naturalized in the West Indies. 






OPLONIA AND PSILANTHELE (ACANTHACEAE) 265 



nema were characterized by " Spangenpollen ", Lindau did not associate it with the 
latter but with Carlowrightia, Anisacanthus and Harpochilus. 

In 1904 Lindau added to his genus Psilanthele a second species, P. grandiflora 
(PI. 42) collected by Ernst Ule {Ule 64g3) in north-eastern Peru, which differed from 
the type-species in having larger flowers, the corolla being about i cm. long, in con- 
densed axillary fascicles. He noted that this had two minute staminodes and 
" pollinis granula typica, fere globosa, c. 38 /z diam. " In the following account, 
Psilanthele is restricted to its type species and P. grandiflora is transferred to Oplonia. 

The extensive collections made by William Harris in the interior of Jamaica 
brought to Lindau's attention a species agreeing in its contracted inflorescence 
with the Peruvian P. grandiflora but having larger flowers, the corolla being nearly 
2 cm. long, although with " pollinis granula iis Graptophyllearum conformia, sub- 
globosa, 46-50 n diam. " Lindau named this P. jamaicensis in 1908. That year 
Harris collected material of an allied plant on the coast of Jamaica which Lindau 
described in 1912 as P. minor (Text-fig. i). He then pointed out that Psilanthele 
resembled Anthacanthus so closely in habit and flower that only the constant lack of 
spines in the former and a difference in pollen separated them : " Psilanthele zeigt 
stets deutlichen Rahmenpollen, Anthacanthus nur Spangenpollen ". A Jamaican 
specimen without spines {Harris 10224) and identical with P. minor but possessing 
" Spangenpollen " he accordingly referred to Anthacanthus. Lindau took this as 
evidence of the over-riding importance of the character of the pollen grain, as opposed 
to general resemblance, in the classification of the Acanthaceae. With equal cogency 
he could have taken it to exemplify the danger of over-emphasis on a slight difference 
in the marking of the pollen grain when countered by resemblances in everything 
else. 

Increase in the number of specimens available for study has increased, not lessened, 
the difficulty of defining taxa within this group. When travelling and collecting 
in Jamaica in 1955-56 (cf . Steam, 1959) I collected specimens of these plants in many 
parts of the island. A low-growing white-flowered species occurring in the parishes 
of St. Thomas and Portland, subsequently identified as Swartz's Justicia acicularis 
(now Oplonia acicularis), was easily recognizable in the field as distinct from the other 
members of the genus ; a taller-growing species, now referred to 0. microphylla, with 
very small leaves and pale lilac flowers found on coastal limestone near Kingston 
has also a distinctive appearance. The large leaves and numerous flowers of 
0. jamaicensis set this apart from the others, as regards which there appeared to be 
no evident pattern of variation with convenient discontinuities. Instead the speci- 
mens collected merely indicated the existence of numerous local populations, probably 
themselves more or less homogeneous and with their own assemblage of minute 
characters but not well differentiated from other local populations. Preparation of 
an account of the Acanthaceae for the continuation of Fawcett and Rendle's Flora of 
Jamaica made it necessary to classify this bewildering and exasperating material. 
Questions of nomenclature then arose, which involved taxonomic decisions ; whether, 
for example, early names based on material from other West Indian islands could be 
applied to Jamaican taxa. In dealing with these it became painfully clear that a 
revision of the group over its whole range from the Bahamas to the Grenadines and 



I 



266 A SURVEY OF THE TROPICAL GENERA 

into Peru could not be avoided. The present paper is the outcome of this survey' 
and is based on about 250 gatherings. In 1900 Lindau based his account on some 60 
gatherings ; Nees von Esenbeck in 1847 had about 12 gatherings. The close resem- 
blance of Jamaican species placed in Opionia and Madagascan species placed in 
Forsythiopsis was quite unexpected and did not become evident until long after the 
completion of the account of the American group. 

TAXIONOMIC CHARACTER 

Lindau (1900) wrote of his " Anthacanthus acicularis " that it was " eine sehr 
variabele Species, sowohl was Behaarung wie auch Blattgrosse betrifft. Dadurch, 
dass die Dornen bald lang bald kurz, bald gerade abstehend oder im Winkel abstehend, 
bald etwas gebogen, bald sehr diinn, bald holzig dick sind, wechselt der Habitus 
ungemein. " Within this he included the plants distinguished by other authors as 
Anthacanthus microphyllus, A. cuneatus, A. lycioides, A. purpurascens and A. tetra- 
stichus, although not, as it happens, any specimens of the true A . acicularisl He thus 
dismissed as having no taxonomic value their diversity in form and length of spines. 
Of his A. spinosus he wrote similarly ; " Ebenso wie vorige Art sehr variabel im 
Habitus. Die Veranderlichkeit erstreckt sich auf die Stacheln und die Blatter ", 
He was left with " fiores + puberuli " as the distinguishing feature of A. acicularis 
and " fiores semper glaberrimi " oi A. spinosus. The characters mentioned above 
cover the major variation within the genus. They may be summarized as follows : 

Habit of Growth. 
Most members of the genus are shrubs up to 2 metres high. The extremes are 
represented by Oplonia nannophylla (PI. 45) of Cuba, a shrublet only 4-8 cm. high 
with internodes 1-3 mm. long, and 0. armata var. pallidior of Jamaica, which may 
grow into a small tree. Possibly each species when living has a distinctive habit. 
As noted above, I found 0. acicularis easily recognizable in Jamaica by its low, much- 
branched growth as well as by its white flowers. 

Pubescence of Branches and Spines. 
With age the shoots of Oplonia lose their original epidermis, this being replaced by a 
thin bark, and become glabrous. As in Thymus {Labiatae) and Epilobium [Ona- 
graceae), the young four-sided stems may be holotrichous, i.e. with hairs on all sides, 
or goniotrichous, i.e. with the two opposite sides immediately below the leaves 
glabrous and the two opposite intervening sides hairy, their position thus alternating 
from node to node. The hairs themselves may be relatively long (up to 0-5 mm. long) 
or much shorter (down to o-i mm. long), dense or sparse, horizontally spreading or 
slightly or distinctly upcurved. 

Variation in the pubescence of shoots is well illustrated by a series of specimens 
belonging to a small-leaved taxon (here included in Oplonia microphylla) which grows 
on the dry limestone over a limited area of south-eastern Jamaica near Kingston 
(Text-fig. 2 J-M) . The type-specimen of A nthacanthus cuneatus Nees [Macfadyen s. n. 
in Herb. Kew) exemplifies one extreme : white spreading hairs up to 0-5 mm. long 
cover densely all four sides of the shoots, while the spines are densely pubescent with 



OPLONIA AND PSILANTHELE (ACANTH ACEAE) 




Fig. 2. Spines and leaves of Oplonia : A,B,C, O. spinosa (Jacq.) Raf. (A, Buck 1851, Haiti ; 
B, Curiiss 133, New Providence ; C, Marie- Victorin et al. 21625, Cuba) ; D.E, O. tetrasticha 
(Griseb.) Stearn var. tetrasticha (D, Proctor 16430, Cuba ; E, isotype, Wright 306J, Cuba) ; 
F,G, O. tetrasticha var. polyece Stearn (F, Clemente 5014, Cuba ; G, holotype, Clemente 
565J, Cuba) ; H,I, O. acicularis (Swartz) Stearn (H, authentic specimen, Swartz, Jamaica; 
I, Stearn 4g8, Jamaica) ; J,K,L,M, O. microphylla (Lam.) Stearn (J, isotype, Richard, 
St. Croix ; K, Howard iogi6, Grenadines ; L, Araque-Molina&^Barkley 22 J 448, Jamaica ; 
M, holotype of Anthacanthus cuneatus, Macfadyen, Jamaica) ; N, O. nannophylla (Urban) 
Stearn (holotype, Ekman 16532, Cuba) ; O, O. armata var. pallidior Stearn {Steam 
863, Jamaica) ; P,Q, O. purpurascens (Griseb.) Stearn (isotypes, Wright 3066, Cuba) ; 
all X i^. Cystoliths are present on all leaves but are indicated only in figs. H and I. 



268 A SURVEY OF THE TROPICAL GENERA 

shorter hairs. Similar abundant pubescence of branches and spines occurs in 
Araque-Molina & Barkley 22 J 448 and Proctor ioig$, both gatherings from near 
localities in Kingston and St. Thomas ; Proctor ggyi and Asprey 20 ^g from St. 
Catherine and Proctor 18268 from St. Thomas have the shoots likewise densely hairy 
but the spines glabrous ; Harris g6o6, Proctor 11436, Steam 66, and Steam 830, all 
from the Long Mountain, Kingston and St. Andrew, have two opposite sides densely 
hairy but the other two sides almost glabrous and the spines glabrous. In Harris 
86og from St. Andrew the hairs are much more abundant on two opposite sides than 
on the intervening sides but the hairs are only o-i to 0-2 mm. long. Thus three types 
of hair covering, which in some genera, e.g. Thymus, would be taken as evidence of 
specific distinctness, here occur on plants of similar habit growing under similar 
conditions in the same region. 

spines. 
The spines of Oplonia (Text-fig. 2 ; PI. 45) are reduced shoots ; hence they arise 
singly and opposite one another in the axils of leaves, each node having but one pair. 
The apparent exception, Grisebach's Anthacanthus bispinosus with two pairs of 
spines at a node, does not belong to Oplonia, its spines being modified leaves, and 
proves to be Barleriola solani folia (L.) Oerstedt, specimens of which may be found in 
herbaria determined as Anthacanthus spinosus (e.g. /. H. Jack 55g6) and A. tetra- 
stichus (e.g. A. Gonzales 6y2). Spines appear to be constantly absent from the 
mesophytic 0. jamaicensis and 0. grandiflora. Among the specimens referred to 
0. armata var. pallidior some have no spines, other have well developed spines ; 
indeed the same bush may bear spiny and spineless shoots, as on Steam 863 from St. 
Elizabeth. Spines are usually well developed in the xerophytic species. They then 
supply some conspicuous diagnostic characters. Thus spines spreading horizontally 
but with a slight downward curve or curved so strongly as to be hooked or deflexed 
characterize 0. spinosa as accepted here. Such spines are found on specimens j 
collected from the Bahamas to the Virgin Islands but not on any Jamaican specimens. 
Very long, straight, usually stout and rigid spines, varying between i cm. and 3-5 cm. 
in length, distinguish 0. tetrasticha of Cuba ; spines more than 1-4 cm. long are un- 
known outside Cuba. Spines very slender for their length, i.e. 8 mm. to 14 mm. lon§ 
but about 0-3 mm. thick, help to give 0. acicularis of Jamaica its distinctive habit.] 
In 0. microphylla they are shorter and more rigid. The development of pairs of 
short subsidiary spines on the ordinary spines is characteristic of 0. tetrasticha var.j 
polyece. The Madagascan species have no spines. 

Leaves. 

The largest leaves of the genus are found in 0. grandiflora (PI. 42) and 0. jamaicensisl 
where they vary on the same plant from 2 cm. to 10 cm. in length and to 3 cm. ori 
more in width. At the other extreme are 0. nannophylla, 0. linifolia and 0. microphyllal 
with leaves only 2-10 mm. long. On many specimens the leaves seem fairly} 
constant in size and shape but on others there may be remarkable diversity. 
Thus a gathering made on St. John, Virgin Islands {Robertson 26), here referred to] 
0. spinosa, has the leaves almost circular or very broadly ovate, the length and] 
breadth being almost the same with the greatest breadth at or below the middle. On] 






OPLONIA AND PSILANTHELE (ACANTHACEAE) 269 

a gathering by A. C. Smith [A. C. Smith 10562) from Virgin Gorda, Virgin Islands, 
also referred to 0. spinosa, some leaves are likewise broadly ovate but others are 
elliptic or narrowly obovate, i.e. proportionally narrower with the greatest breadth 
at or above the middle. Such specimens connect with the narrowly elliptic and 
oblanceolate leaves of 0. spinosa on specimens collected in Haiti and the Bahamas. 
The apex of the leaf is acute to acuminate in 0. acuminata and 0. grandiflora. The 
Madagascan 0. linifolia has linear leaves. The others have the apex itself rounded 
and usually slightly or distinctly indented, sometimes with a distinct mucro. 

The veining (PI. 43b) is evident in some specimens, faint or invisible in others, and 
seemed at first to offer characters of possible taxonomic value. Using the clearing 
technique described by Arnott (1959), preparations were made of leaves from a 
diversity of plants, these inserted in a microfiche viewer, and their veining then 
drawn on a vastly enlarged scale. It proved to be brachydromous in all, i.e. with 
about 6 to II lateral veins on each side of the midrib running slightly obliquely 
almost to the margin, then bending abruptly forward and joining the bend of the 
lateral vein immediately above, thus making a closed series of loops more or less 
parallel to the margin. Variation in the number of veins and their angle of diverg- 
ence from the midrib was slight. 

Number of Flowers. 
The inflorescence of Opionia occasionally develops into a short lateral raceme but is 
usually reduced to a sessile axillary fascicle, which may bear up to 15 flowers in the 
large-leaved 0. jamaicensis but is commonly only one- or two-flowered in species with 
leaves less than 2-5 cm. long. 

Pedicels. 
The pedicels are mostly much longer than the calyx and up to i cm. long. In 
0. purpurascens, however, the flowers are almost sessile with pedicels 1-1-5 i^^- lo^^g 
and thus shorter than the 3-4 mm. -long calyx. Specimens of 0. microphylla some- 
times have pedicels as short, but longer pedicels are usually to be found on the same 
plant. They are usually glabrous, but occasional individuals have hairs on the 
pedicels. 

Calyx. 
The calyx is parted almost to the base into 5 narrow segments which are linear in 
0. grandiflora but very narrowly triangular (subulate) in the others. The calyx is 
usually about ^-5 mm. long. Its length does not seem correlated with other char- 
acters. In specimens referred to 0. armata var. pallidior it varies from 2 mm. 
{Howard, Proctor & Steam 14740, Norman 100) to 5 mm. {Steam 863). 

Corolla 

The corolla (Text-fig. i, 3) is surprisingly uniform in size and form and consists of a 
slender tube slightly expanded above the insertion of the stamens on its upper part 
and a spreading two-lipped limb with two ascending more or less fused upper lobes 
and three descending distinct lower lobes. Tube and limb are almost equal in length. 
In 0. grandiflora, so inappropriately named, the tube is only 4-4-5 mm. long, in 0. 
acicularis 4-6-5 mm., in the others 6-10 mm. long. Although Lindau clutched at the 



270 



A SURVEY OF THE TROPICAL GENERA 



presence or absence of hairs on the corolla in despair of finding any other character 
for specific distinction, this seems of trivial value. The hairs are very small and' 
are present or absent without correlation with any other characters. 

The red colour of the corolla of 0. armata var. armata readily distinguishes it in a 
living state from all other members of the genus. In these the corolla is mostly a 
bluish lilac or bright purple having at the base of the lower lip a pale area marked 
with dots or streaks of purple, which presumably serve as nectar guides. In 
0. acicularis the corolla is normally white. 

Length of Stamens and Style 

In some flowers of Opionia the stamens project conspicuously from the tube of the 
corolla, in others they scarcely reach beyond its mouth (Text-fig. 3). This diversity 




Fig. 3. Floral dimorphism in Opionia armata (Swartz) Stearn ; flower cut lengthwise 
A.B, O. armata var. pallidior Stearn (A, Adams 1114J, Round Hill, Clarendon, Jamaica ; 
B, Adams 11143, same locality) ; C,D, O. armata var. armata (C, Adams 11062, between 
Aenon Town and McKay, Clarendon, Jamaica ; D, Adams 11060, same locality). Flower, 
X 2, leaves x i. 

seemed at first to offer welcome specific characters but undoubtedly expresses flora 
dimorphism, merely separating individuals within the same population, such asj 
occurs in many gamopetalous families, notably the Rubiaceae (cf. Burck, 1883-4 
Bahadur, 1963), although not previously recorded for the family Acanthaceae. ThusJ 






OPLONIA AND PSILANTHELE (ACANTHACEAE) 271 



in the specimens of south-eastern Jamaica mentioned above as referred to 0. micro- 
phylla and agreeing closely in habit, some have short stamens and long styles, others 
long stamens and shorter styles. For example. Proctor 18268 from St. Thomas, 
Harris 86og from Kingston and St. Andrew, Asprey 20 5g and Proctor ggyi from St. 
Catherine have the filaments about i-2-i-5 mm. long, the style 7-9 mm. long. 
Araque-Molina & Barkley 22J448 from St. Thomas, Proctor 11436, Stearn 66 and 
Steam 830 from Kingston and St. Andrew have the filaments about 5-5-5 mm. long, 
the style 4-5-5 mm. long. Both types occur on the Long Mountain near Kingston. 
Among specimens of 0. armata var. pallidior, Stearn J26 has the filaments 1-5 mm. 
long, the style 10 mm. long, but Proctor 166^^ has the filaments 5-5 mm. long, the 
style 6 mm. long. Likewise, of specimens of 0. armata var. pallidior collected by 
Dr. C. D. Adams on Round Hill, Clarendon on 13 May 1962, Adams 11147 has long- 
exserted stamens and a short style, Adams 11143 short stamens and a long-exserted 
style (Text-fig. 3, A,B). Similar dimorphism is found in 0. armata var. armata 
collected by Adams between Aenon Town and McKay, Clarendon, Adams 11062 
(with light crimson corolla) having long-exserted stamens and a short style, Adams 
11060 (with rich- violet corolla), short stamens and a long-exserted style (Text-fig. 
3, C, D). 

The same floral dimorphism is evident in Cuban material of 0. spinosa all from the 
Moa region of Oriente province. Thus Victorin, Clemente & Alain 21621 and 2162^ 
have the filaments about 2 mm. long, the style 10 mm. long, but in Clemente 366^ and 
3646 the filaments are about 4-5 mm. long, the style about 5-6-5 mm. long. Such 
floral dimorphism probably occurs throughout the genus ; several specimens agreeing 
in other respects and from the same district or, better, the same locality are needed 
to demonstrate this. Its development is of great evolutionary interest. Its presence 
in Opionia raises the question whether it may not have been overlooked elsewhere 
in the family and have thus led to unjustified separation of long-stamened and short- 
stamened forms as different species, although, as Darwin (1877 : 136)* pointed out, 
in the Rubiaceae " this character, which is not even of specific value in the hetero- 
styled species, is often of generic value in other members of the family ". It also 
occurs in Madagascan specimens formerly placed in the genus Forsythiopsis. Thus 
in a gathering of 0. minor {Humbert 2^06) one specimen has the filaments about 
2.5 mm. long and the style 7 mm. long, another has the filaments about 5 mm. long 
and the style 5 mm. long. A specimen of 0. vincoides collected near Fort Dauphin 
by Cloisel {Cloisel 2og) has the filaments about 3 mm. long, the style about 10 mm. 
long, while another, also from Fort Dauphin, collected by Scott Elliott {Scott Elliott 
2586) has the filaments about 6 mm. long, the style 6-8 mm. Thus heterostylous 
floral dimorphism probably occurs throughout the genus. 

Pollen 

Lindau attributed " SpangenpoUen " to Opionia and " RahmenpoUen " to 
Psilanthele, using this as a generic distinction. Actually the difference between 
" RahmenpoUen " and " SpangenpoUen " is by no means clear-cut. The pollen 
grains of typical species of Opionia such as 0. microphylla (Text-fig. 4) and 0. armata 

* The " Faramea" described by Darwin has been discussed by H. G. Baker, " Pollen dimorphism in 
the Rubiaceae ", Evolution lo : 23-31 (1956), 340 (1956) and proves to be Rudgea jasminoides Muell.-Arg. 



272 



A SURVEY OF THE TROPICAL GENERA 



var. pallidior (PI. 38) are spheroid or prolate, 30-50 ^ m. in diameter, with three germ 
pores each in a longitudinal furrow (mesocolpium) reaching almost to the pole, the 
grain thus being primarily three-furrowed (tricolpate), but with a shorter subsidiary 
longitudinal furrow (pseudocolpus) on each side of the main furrow ; all nine furrows 
are separate and do not join. They are " SpangenpoUen ", as figured by Lindaui 
(1893 : 38, fig. I L, t.2 fig. 62 ; 1900 : 173, fig. 12, 13), since the subsidiary furrows! 
remain distinct and do not continue towards the poles and there fuse. In " Rah- 1 
menpollen ", as figured by Lindau (1893 : fig. i O, t.i fig. 49-51 ; 1900 : 173, fig. 10, 
11) and attributed by him to Psilanthele, the two adjacent subsidiary furrows meetj 
near the poles, thus joining into three pairs. If these furrows converge but do not 
quite touch, then presumably the grain belongs to the "SpangenpoUen" type] 
(cf . PI. 38, 40) ; if they join, then it belongs to the " RahmenpoUen " type (cf. PI. 39). 




Fig. 4. Pollen grains (" SpangenpoUen ") of Oplonia microphylla (Lam.) Steam {Steam 66 J 
drawing based on Stereoscan photograph), x 700. 

This distinction is not always easy to observe even on grains stained and magnifiec 
X 1500, yet it is the only one, following Lindau, by which the Jamaican species 
referred to Psilanthele can be distinguished generically from Anthacanthus, i.e.| 
oplonia. Actually there is no difference in pollen grains between Opionia anc 
Psilanthele. Photographs made with the scanning electron microscope (Cambridge 
Instrument Co. Stereoscan), at magnifications between x 500 and x 5,700, of poller 
grains of Oplonia microphylla {Steam 66), 0. armata var. armata {Proctor 21407) \ 
0. armata var. pallidior {Steam 726), 0. acicularis {Britton 28g4), 0. jamaicensis 
{Harris 8g68), 0. grandiflora {Ule 64g3), the Madagascan 0. vincoides {Baron ig37y 
McWhirter & Capuron 184) and Psilanthele eggersii {Eggers I5i2g) show them to be 
essentially uniform in form, although varying in size, with their subsidiary furrows 
sometimes joining but usually not quite reaching each other. The type species of 
Psilanthele {P. eggersii of Ecuador) differs from Lindau's Peruvian and Jamaican! 
species not in pollen grains but in its loose inflorescence and very small flowers withi 
the two loculi (thecae) of the anther separated a little, not absolutely parallel and! 
contiguous as in Oplonia. Hence the genus Psilanthele is here restricted to P.\ 
eggersii and the other species are transferred to Oplonia. 



OPLONIA AND PSILANTHELE (ACANTH ACEAE) 273 

Variation in size of pollen within a species can be illustrated from Jamaican 
material of 0. micro-phylla, measurements taken by Dr. S. Sengupta from lo grains 
each of 8 specimens being as follows : 

1. Proctor 18268 : diameter range 33-3 ^m-36-o ptm. ; spheroid. 

2. Proctor 24186 : diameter range 34-2 /im-36-o nm ; spheroid. 

3. Steam 66 : diameter range 31-5 ^m-32-4 /im ; prolate, with length range 
43-2 fim-so-4 fim. 

4. Webster, Ellis & Miller 8364 : diameter range 28-8 /im-29-7 fim ; prolate with 
length range 45-0 fmi-4g-^ fim. 

5. Harris g6o6 : diameter range 28-8 fim-zg-y jum ; prolate, with length range 
40-5 /im-43-2 ^m. 

6. Araque-Molina & Barkley 22J448 : diameter range 32-4^m-36-0)um ; spheroid. 

7. Steam 830 : diameter range 35-2 /im-37-8 jum ; prolate-spheroid, with length 
range 387 ^m-40'5 fxm.. 

8. Proctor ggyi : diameter range 27-0 /im-28-8 fim ; prolate-spheroid, with length 
range 28-8 /im-32-4 ixm. 

TAXONOMIC PROCEDURE 

As noted above, the plants of this group show diversity in habit, in length, form 
and direction of spines, hairiness of branches, form and size of leaves, number of 
flowers at an axil, length of pedicel, length of calyx, colour and pubescence of corolla, 
and length of filaments and style, without there being much readily evident corre- 
lation between the characters of one organ and those of another. The difficulty of 
perceiving associations is increased by the smallness of the organs themselves, which 
require examination one by one with a lens ; each character thus presents itself to the 
mind as a separate entity not linked with others. Lindau's treatment of 1900 brought 
together under one species a multitude of very unlike specimens while at the same 
time referring almost identical specimens to different species. Dissatisfied by this 
and by the evident confusion, I decided to ignore for the time being all literature 
about the group and all determinations on the herbarium sheets, and instead to begin 
anew by treating each specimen as if it were a new and independent taxon, tabulating 
its characters and then trying to ascertain how these were correlated with others. 
This proved a laborious and time-consuming procedure, despite the independent use 
of three complementary techniques, i.e. punched cards, metroglyphs and correlation 
coefficients. Firstly, characters of potential utility were recorded on marginally 
punched cards. These made it possible to check the distribution of characters and 
to test likely combinations, usually with negative results ! Secondly, a chart was made 
by expressing a few salient characters of each specimen (e.g. the size and shape of the 
leaves, the type of pubescence and the colour of the corolla) as pictorialized scatter 
diagrams or metroglyphs (cf. Text-fig. 7 B), following Edgar Anderson (1952 ; 1957 ; 
see also Hatheway, 1962.) Thirdly, with the help of Dr. P. H. A. Sneath, a range of 
33 character states on 68 specimens were coded for computer-aided single-linkage 
cluster analysis ; Sneath then calculated their correlation coefficients (cf. Sneath, 
1962) and produced a dendrogram indicating overall similarity in the characters taken. 
Those available in Oplonia fell far below the postulated minimum of forty characters 



274 A SURVEY OF THE TROPICAL GENERA 

and some unexpected associations resulted. Metroglyphs seemed to provide the 
best method of grasping the essentials of this complex taxonomic situation, punched 
cards a means of quickly retrieving the information, correlation coefficients a means 
of grouping objectively the specimens or taxa. In fact none of these methods gave a 
satisfying arrangement by itself. They resulted, however, in bringing together piles 
of specimens having many characters in common, which it had been difficult or 
impossible to associate by simple observation, since there were too many inconspicuous 
variables concerning too many specimens for them all to be kept in mind at the same 
time. It then became possible to take more or less homogeneous clusters of specimens 
and associate other specimens with them, later uniting these groups with other 
groups as the discontinuities between the groups were obliterated or bridged. Be- 
fore this process merged the twenty or so provisional West Indian taxa into one, the 
dominance of an easily observable feature in a particular group and its absence from 
others, e.g. very small leaves, hooked or decurved spines, very long spines, numerous 
flowers in a fascicle, or red corolla, became evident. Most of the material could then 
be placed into groups characterized by these features ; ambiguous specimens were 
referred to them on a basis of general resemblance to already placed specimens from 
the same island or district. The result has led me to accept nine generally recogniz- 
able but not always clearly distinguishable American groups treated below as 
species. 

TAXONOMETRIC SURVEY 

The most difficult task when classifying apparently rather uniform biological 
material, such as that of Oplonia, wherein most distinctions lie in a diversity of 
minute characters neither simultaneously observable nor easily correlated, is that of 
putting the specimens into small relatively homogeneous groups which can later be 
joined into larger groups as defining attributes become evident ; obviously then, the 
study of such material would be eased by the use of an efficient computer-aided 
method of assembling the specimens into successively united groups (clusters) 
characterized at first by high overall resemblance between the members of each 
group, a resemblance which would decrease progressively as more groups were 
united. Hence it seemed worthwhile, as a guide for future procedure in dealing with 
intractable groups like Oplonia, to apply such methods to an assemblage of specimens 
of Oplonia. 

Availability of computers and exercise of mathematical ingenuity have now led to 
the creation of possible numerical methods bewildering in their number and claims 
(cf. Johnson, 1968, for comment and extensive bibliography ; also Cole, 1969). 
Assessment of their respective merits for a given task must depend upon practical 
experience. They need to be tested by application to group upon group (cf . Stearn, 
1969). 

The following observations relate to a method of cluster analysis based on graph 
theory, to which Professor D. J. Rogers (then at the New York Botanical Garden, 
now at the University of Colorado) drew my attention in 1964, after the present study, 
mostly done in 1961-62, had been virtually completed. This method has since been 
expounded mathematically by Estabrook (1966) and applied by Wirth, Estabrook 



» 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 275 



and Rogers (1966) to Orchidaceae subtribe Oncidiinae, by Irwin and Rogers (1967) to 
Cassia sect. Apoucouita and by Prance, Rogers and White (1969) to Chrysobalanaceae 
tribe Hirtellae. It groups objects into clusters, defined by similarity of their mem- 
bers, with stage-by-stage reduction in the degree of similarity necessary for objects 
to be placed in the same cluster. It is a process wherein hitherto unclustered objects 
are progressively admitted into earlier-formed clusters through linkage by resem- 
blance to objects already in the cluster. Thus small clusters become linked to larger 
clusters ; as a cluster enlarges through incorporation of other clusters, so the amount 
of similarity between all members of the resulting cluster lessens at each stage until 
ultimately the objects all come together into one large cluster. The computer print- 
out for the whole process records it as a series of stages of clustering. The taxo- 
nomist must decide which of the intermediate stages, when there exist several or 
many clusters of specimens, have the most relevance to his needs. 

The first task, as in other taxonometric procedures, consists of listing all the 
available characters and their states differing from specimen to specimen ; these are 
then tabulated for each specimen being studied. Comparison by the computer pairs 
the specimens according to their mutual similarity. This stage, the first partition, 
divided the material into the basic classes. The principles governing procedure, as 
stated by Estabrook (1966), are : 

" (i) a biological classification for a collection of objects is a series of partitions for 
this collection, classes under later partitions consisting wholly of classes under earlier 
partitions. (For example, genera consist wholly of species.) 

" (2) For any given partition in this series, two similar objects should not be placed 
in different classes (this assumes that no knowledge of evolution is available) . 

" (3) The classes of a given partition should be isolated from one another ; that is, 
there should be some phenotypic discontinuity between members of different classes. " 

As the original clusters (disjoint partitions) merge into the ultimate single cluster 
(conjoint partition), by coming together at given degrees of lessening similarity, the 
greater the overall resemblance between specimens the earlier will be the stage at 
which they are linked into a common cluster ; conversely, the greater their divergence 
so much the later will be their linkage. Part of the task of preparing a taxonomic 
revision of a group has always been a search for such resemblances and divergences, 
followed by hierarchic ranking of the sub-groups detected. A machine technique 
quickly displaying these relations can thus significantly lessen the labour of research ; 
it may even indicate correlations of characters which otherwise would not have been 
perceived. 

To test this computer-aided hierarchical clustering technique based on graph 
theory, the characters, 26 in all, of 35 specimens of Oplonia representing the geo- 
graphical and morphological extremes of the group as well as numerous intermediate 
states were tabulated by me and then computed (using Taximetric Similarity Graph- 
Clustering programme on CDC 1604 computer at New York)* by Rogers, who con- 
verted the computer print-out into a series of diagrams (subgraphs), each represent- 

* This study was run on a Control Data 1604 under the direction of D. J. Rogers of the Taximetrics 
Laboratory then located at the New York Botanical Garden but now at the University of Colorado, 
Boulder, Colorado, U.S.A. The programmes are operative on IBM 7000 series + 360/75 and CDC 6400. 



276 



A SURVEY OF THE TROPICAL GENERA 



ing one partition of 21 stages (levels) of clustering. These specimens (to which the 
scientific name now adopted has been added) were as follows : 



I. Britton28yy ; 


Jamaica ; \ 
Jamaica ; 


0. acicularis ■ 


2, Steam 550 ; 


0. acicularis 


3. Britton 28g4 ; 


Jamaica ; 


0. acicularis 


4. Proctor 8433 ; 


Jamaica ; 


0. jamaicensis 


5. Steam 454 ; 


Jamaica ; 


0. jamaicensis 


6. Harris 8g68 ; 


Jamaica ; 


0. jamaicensis (isotype) 


7. Proctor 16430 ; 


Cuba ; 


0. tetrasticha var. tetrasticha 


8. Leon 3326 ; 


Cuba ; 


0. tetrasticha var. tetrasticha 


9. Curtiss 133 ; 


Bahamas ; 


0. spinosa 


10. Robertson 26 ; 


Virgin Islands ; 


0. spinosa 


II. Steam y26 ; 


Jamaica ; 


0. armata var. pallidior 


12. Proctor 2i4oy ; 


Jamaica ; 


0. armata var. armata 


13. Box ysy ; 


Antigua ; 


0. microphylla 


14. Howard iogi6 ; 


Petit St. Vincent 


; 0. microphylla 


15. Araque-MoUna & Barkley 






22] 448 ; 


Jamaica ; 


0. microphylla 


16. Proctor ggy I ; 


Jamaica ; 


0. microphylla 


17. Steam 830 ; 


Jamaica ; 


0. microphylla 


18. Norman 100 ; 


Jamaica ; 


0. armata var. pallidior i 
0. armata var. pallidior 


19. Steam 863 ; 


Jamaica ; 


20. Ekman 15042 ; 


Cuba ; 


? 0. tetrasticha var. tetrasticha 


21. Marie-Victorin, Clemente 






& Alain 21625 ) 


Cuba ; 


0. spinosa 


22. Leonard & Leonard I36y4 ; 


Hispaniola ; 


0. spinosa 


23. Valeur y24 ; 


Hispaniola ; 


0. microphylla 


24. Ekman 16534 '> 


Cuba ; 


0. nannophylla (type) 


25. Clemente 56 5y ; 


Cuba ; 


0. tetrasticha var. polyece 

(type) 


26. jBwc/f j55j ; 


Hispaniola ; 


0. spinosa 


27. Proctor 10236 ; 


Jamaica ; 


0. armata var. armata 


28. Schomburgk loy ; 


Hispaniola ; 


0. spinosa 


29. Steam 486 ; 


Jamaica ; 


0. armata var. armata ^ 


30. Wright 3o6y ; 


Cuba ; 


0. tetrasticha var. tetrasticha 
(isotype) 


31. Clemente 5014 ; 


Cuba ; 


0. tetrasticha var. polyece 


32. 5/eam ^55 ; 


Jamaica ; 


0. armata var. armata 


33. Proctor 8830 ; 


Turks & Caicos 






Islands ; 


0. spinosa 


34. Steam 4g8 ; 


Jamaica ; 


0. acicularis 


35. C//5 649J ; 


Peru ; 


0. grandifiora (isotype) 



Study of Rogers's series of diagrams (subgraphs) in association with the specimens 
themselves revealed particularly the utility of the middle stages (levels) of clustering 






OPLONIA AND PSILANTHELE (ACANTHACEAE) 



277 



as an aid to perceiving and distinguishing the main components of the group. 
Text-fig. 5, for example, represents the clustering reached at stages 6 and 12. The 
first partition (stage i) made according to a very high degree of resemblance brought 
together the almost identical Jamaican specimens (Nos. 15, 16, 17) now referred to 0. 
microphylla. By stage 6 Jamaican specimens (Nos. 11, 12 and 18, 19) now referred 
to 0. armata had been placed in two clusters, Lesser Antillean specimens (Nos. 13, 
14) referred to 0. microphylla in another, Cuban specimens (Nos. 7, 8) referred to 0. 



STAQE 6 



tCD — H^ 

M [13] [14] M 



M 15 

/ \ 





A 18 



19 A 



aIZI — B'^ 

^0 — Hj 



STAGE 12 



T 30 




26 S 




T? 


S 




T 7 


20 


21 28 


T 8 








T [25 


. 


S 22 9 




M 






A 19 


16 23 M 

V W V 


A 

a|1 


— 1 17 13 M 


M 




A II 


A 27 
J \ 
12 A 




J 




J |_6 


4 




Ll ^ 



Fig. 5. Stages 6 and 12 of 1-22 sequence in cluster analysis of 35 specimens of Oplonia 
(listed by numbers on p. 276) by graph-theory method as computed by D. J. Rogers ; 
A, specimens classified as O. armata ; J, as O. jamaicensis ; M as O. microphylla ; S as 
0. spinosa ; T as O. tetrasticha. 

tetrasticha and Jamaican specimens (Nos. 4, 5) referred to 0. jamaicensis in yet others. 
By stage 12 the grouping and linking have gone far enough to make a significant 
arrangement of most of the specimens. Thus Nos. 13, 14, 15, 16, 17, and 23 when 
put side by side agree in having very small leaves and short ascending spines and 
constitute a taxon for which the earliest epithet is microphylla ; Nos. 7, 8, 20, 25 and 



278 



A SURVEY OF THE TROPICAL GENERA 



I 



30, all from Cuba, with their long straight horizontal spines form another group, for 
which the epithet is tetrasticha ; Nos. 9, 21, 22, 26, and 28 with their mostly curved 
horizontal or defiexed spines likewise form a group, to be called spinosa ; Nos. 4, 5 
and 6, all from Jamaica, with large leaves and many flowers represent the easily 
distinguishable jamaicensis. Among the specimens unclustered at this stage. No. 24 
{nannophylla) , Nos. i, 2, 3 and 34 {acicularis) and 35 {grandiflora) strike the eye as 
well distinct from the others, while Nos. 32 [armata), 10 {spinosa) and 33 [armata] 
show resemblance to some specimens already clustered. The last specimen to enter 
was No. 35 {grandiflora), from Peru, which accords with its morphological and 
geographical distinctness. The logical coherence of this grouping made me regret 
that the technique used had not been available much earlier ; it would have provided 
a means of grouping fairly quickly the specimens into the taxa ultimately adopted 
without so much time-consuming search for correlations. The computer run dealing 
with this material took 3 minutes 47 seconds, an infinitesimal amount of time com- 
pared with that of ascertaining and tabulating data and that of expressing and con- 
sidering the results. The method commends itself for its graphic portrayal of classi- 
fication as a stage-by-stage process and corresponds closely enough in its working to 
the mental acts of a working taxonomist to be able to serve as a time-saving substi- 
tute for some of them. Thus in dealing with a large confused group it is convenient 
to pick out various types and use these as focal points to which specimens can be 
referred by resemblance, thereby creating assemblages from which concepts of 
taxa can take shape and definitions be formulated. This computer method creates 
such assemblages in the same typological manner, presents them to the eye pictorially, 
and thus prepares the way for defining their circumscription and ascertaining the 
characters by which they can be distinguished. 

The reasonableness of the classifications of genera (cf. Sims, 1966) and species 
(cf. Wilkinson, 1967 ; Stearn, 1969) produced by computation using the computer 
programme CLASP indicated the desirability of testing it as a means of classifying 
individual specimens, i.e. of arranging them into groups to be treated as species or 
subspecies, by taking each specimen as a separate operational taxonomic unit (OTU) 
for association with other specimens on a basis of overall shared characters, as had 
been done using the different technique outlined above. Accordingly essentially the 
same characters (24) of the same specimens (35, plus one of 0. purpurascens from 
Cuba earlier inadvertently omitted) were coded and tabulated for computation as 
follows : 



Two-state (Binary) 

Habit : tall ; low. 

Internodes : 5 mm. long or less ; more than 5 mm. long. 

Hairs of stem : present ; absent. 

Hairs of stem : short ; long. 

Hairs of stem : on 2 sides ; on 4 sides. 

spines : present ; absent. 

spines : short ; long. 

spines : fine ; stout. 



f 



9- 


spines : 


10. 


Leaf-base : 


II. 


Leaf-apex : 


12. 


Leaf-apex : 


13- 


Flowers : 


14. 


Pedicel : 


15- 


Pedicel : 


16. 


Sepals : 


17- 


Corolla : 


18. 


Corolla-tube 


19. 


Spines : 


20. 


Leaf : 


21. 


Corolla : 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 



simple ; branched. 

cuneate ; rounded. 

emarginate ; not emarginate. 

mucronate ; not mucronate. 

I or 2 ; more than 2. 

shorter than calyx ; longer than calyx. 

pubescent ; glabrous. 

subulate ; linear. 

pubescent outside ; glabrous. 

to 4-5 mm, ; more than 4-5 mm. 

Multi-state (Qualitative) 
deflexed or curved ; horizontal ; ascending. 



279 



broadest below middle 
white : blue : red. 



at middle : above middle. 



Quantitative 

22. Leaf length, 

maximum ; less than 12 mm. ; 12-25 ^"i. ; more than 25 mm. 

23. Leaf breadth, 

maximum : less than 6 mm. ; 6-18 mm. ; more than 18 mm. 

24. Sepal length : 2-2-5 ^^'^- ', 2-5-4 ^^i^- '» more than 4 mm. 



A few specimens, being incomplete, did not provide all the information desired and 
these deficiencies, it was understood, had later to be kept in mind since they might 
affect not only the subsequent placing of such specimens but also the linkage of 
others ; thus Nos. 20 and 24 lacked flowers. 

Computation of this data was made by Mr. G. J. S. Ross at the Rothamsted 
Experimental Station, Harpenden, Herts, using an Orion electronic computer. The 
resulting classification, as expressed in a dendrogram (cf . Text-fig. 6) and a minimum 
spanning tree (cf. Text-fig. 7), agreed in general with that obtained by other means 
and undoubtedly would have been very helpful if initially available. The anomalous 
placing, however, of some specimens led to re-examination of their coding. This 
showed that a few attributes had been originally defined somewhat ambiguously or 
too arbitrarily ; thus some spines intermediate in length between the undoubtedly 
long spines of 0. tetrasticha and the undoubtedly short spines of 0. microphylla had 
been classified as " long " for some specimens but " short " for others, thereby 
causing inconsistency and so making non-valid or exaggerated distinctions ; more- 
over a few attributes, it must be confessed, had been erroneously coded. Since 
variability within a population may not be evident from the relatively few specimens 
available in herbaria for some groups, characters were taken, from the individual 
specimens, which a priori might have been of taxonomic value but which are now 
known to vary from individual to individual within a population, e.g. the presence 
of hairs on two or four sides of the stem, on the pedicels and on the outside of the 



28o 



A SURVEY OF THE TROPICAL GENERA 



corolla. In a small sample, using few characters, these may have a distorting effect 
and, when the aim is to define populations, would seem best ignored. In the dendro- 
gram a specimen (No. 6) of the very distinct 0. jamaicensis was not associated as 
closely with Nos. 4 and 5 as it should have been, simply because it had pubescent 
instead of glabrous shoots and flowers. The anomalous placing of specimens caused 
by the unsatisfactory coding of a few characters indicates the closeness of affinity 



70 



80 

— r- 



90 



-^ 



q 



^ 



35 


grandiflora 


6 


jamaicensis 


5 


jamaicensis 


4 


jamaicensis 


32 


armata 


23 


microphylla 


11 


armata 


27 


armata 


10 


spinosa 


31 


tetrasticha 


25 


tetrasticha 


8 


tetrasticha 


22 


spinosa 


16 


microphylla 


15 


microphylla 


17 


microphylla 


14 


microphylla 


13 


microphylla 


33 


spmosa 


9 


spinosa 


28 


spmosa 


26 


spinosa 


21 


spmosa 


20 


tetrasticha? 


36 


purpurascens 


24 


nannophylla 


19 


armata 


18 


ar mata 


29 


ar mata 


12 


armata 


30 


tetrasticha 


7 


tetrasticha 


34 


acicularis 


3 


acicularis 


2 


acicularis 


1 


acicularis 



Fig. 6. Dendrogram showing overall phenetic similarity of 36 specimens of Oplonia (listed 
by numbers on p. 276) as computed from a matrix of 24 characters by G. J. S. Ross. 



I 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 



281 




Ac 34 



Fig. 7. (A). Minimum spanning tree uniting 36 specimens of Oplonia (listed by numbers 
on p. 276), the lengths of segments indicating phenetic distances between specimens as 
computed from a matrix of 24 characters by G. J. S. Ross ; A, specimens classified as 
O. armata ; Ac as O. acicularis ; G. as O. grandiflora ; J as O.jamaicensis ; M as O. micro- 
phylla ; N as O. nannophylla ; P as O. purpurascens ; S as O. spinosa ; T as O. tetrasticha. 



31 / 

A — Ds 



27 



241 1 36 20 

V- - -V - - nn 



20 

n 



25 



-v^ 



29 






I 









.i 



34 



/ I 



y26 y22 VA 



A'al 



□ ,i 



r9'^' 



(B). Part of above minimum spanning tree converted into pictorialized scatter dia- 
gram ; A leaves broadest below middle, D at middle, v above middle ; 1 upward, spines 
ascending, | downward, spines downwards curved, — spines horizontal ; • flowers white. 



282 A SURVEY OF THE TROPICAL GENERA 

within the group. It also serves as a warning of the narrow margin of error in treat- 
ing closely knit groups and the need for care in defining attributes so that essentially 
similar ones are similarly coded ; inconsistencies in these as well as slight errors of 
observation or recording may have their effect disproportionately magnified through 
lack of compensating other characters within a low range of characters. Taking, 
however, the dendrogram (Fig. 6) as a whole, it emphasizes the distinctness of the 
Peruvian 0. grandiflora (No. 35) from the Caribbean group and it brings together 
enough specimens of 0. acicularis (Nos. i, 2, 3, 34), of 0. armata (Nos. 12, 18, 19, 29), 
of 0. jamaicensis (Nos. 4, 5, 6), of 0. microphylla (Nos. 14, 15, 16, 17), of 0. spinosa 
(Nos. 9, 21, 26, 28, 33), and of 0. tetrasticha (Nos. 8, 25, 31, and 7, 30) to form a basis 
for classification, though it is not acceptable as a whole without re-allocation of 
some specimens. 

A minimum spanning tree (cf . Gower & Ross, 1969 ; Ross in Cole, 1969) constructed 
from the same data provides alternative pictorial expression of phenetic linkage and 
distance (Text-fig. 7) when, as here, principal co-ordinate analysis does not help very 
much because the first five latent roots are nearly equal and a plain vector diagram 
(Text-fig. 8) consequently has no clearly informative pattern and gives a confused 
impression. It should be noted that, in a minimum spanning tree, the linkages be- 
tween units and the distances separating them along the branches are significant, not 
the direction and relative spacing of branches which are simply a matter of conven- 
ience in drawing the diagram. Thus, in Text-fig. 7A, Nos. i, 2, 3 and 34 are closely 
and directly linked but their proximity to Nos. 36 and 22 is purely coincidental and 
has no significance, because these are only distantly and indirectly linked, i.e. 
through No. 12. Here 0. grandiflora (No. 35) is again given an isolated position. 
0. jamaicensis (Nos. 4, 5 and 6) links on to 0. spinosa (Nos. 22, 26, 21, 33, 9) and so to 
0. armata (Nos. 19, 12, 29, 27, 18, 11, 32) which has a central pivotal place in the 
genus and from which some others such as 0. acicularis (Nos. i, 2, 3, 34), 0. tetrasticha 
(Nos. 7, 30, 25, 31, 8) and 0. microphylla (Nos. 17, 13, 14, 16, 15) as well as 0. nanno- 
phylla (No. 24) and 0. purpurascens (No. 30) could be derived. Anomalously placed 
specimens are No. 20 (incomplete, now referred to 0. tetrasticha with doubt), No. 28 
(now referred to 0. spinosa), No. 10 (now referred to 0. spinosa) and No. 23 (now 
Teierred to . microphylla) . 

A minimum spanning tree or principal co-ordinate analysis graph can be converted 
into a pictorialized scatter diagram more helpful for correlations of characters simply 
by adding to each position representing a specimen certain symbols indicating 
character states likely to have taxonomic importance, e.g. a triangle for leaves 
broadest below the middle, an inverted triangle for those broadest above the middle, 
a red mark for red flowers, a blue one for blue, lilac or purple flowers, a dot for white 
ones, small symbols for small leaves, larger for large leaves, etc. (cf. Text-fig. 7B). 

To utilize the results of these computer-produced diagrams in practical taxonomy 
the simplest procedure is to lay out the specimens on a large table in a sequence or 
disposition corresponding to the diagrams and then by scanning and study to find 
the major characters diagnostic of the groups and the discontinuities between them. 
Marked discontinuities are almost nonexistent in Oplonia but the minimum spanning 
tree (Text-fig. 7A) can then be seen to have eight major elements. Excluding No. 35 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 



283 



{0. grandiflora) and Nos. 5, 4, 6 {O.jamaicensis) on account of their very large leaves 
and No. 24 {0. nannophylla) for its condensed growth, the most evident common 
denominator for the specimens assembled in each other group is the character of the 
spine, even though in coding this accounted for only 4 of the attributes. Using spine 
characters for diagnosis and now considering also the provenance of the specimens, 



T+3 



'30 



• 16 



•17 



•7 



25 
•14 •12 .10 • 



-5 



4ll^-H K 



32 
28« 



»36 



31 23 

• • 

»8 • • 
20 13 



>18 



19 



►24 



29 

•3 #34 



*5 



•1 



33< 



•11 



•21 



•22 
•27 



•26 



35 « 



6^ 



59 



.. 94 



J- -6 

Fig. 8 Vector diagram showing lack of easily discernible patterns of correlated variation 
in 36 specimens of Oplonia (listed by numbers on p. 276) as computed from a matrix of 24 
characters by G. J. S. Ross ; the vertical co-ordinates are the vectors corresponding to the 
first latent roots, the horizontal co-ordinates those of the second latent roots, of the trans- 
formed matrix. 



No. 28 from Hispaniola with downward-curved spines and No. 10 from the Virgin 
Islands with some spines likewise downward-curved fit into the spinosa group and 
No. 23 from Hispaniola with fine ascending spines fits into the microphylla group. 
The same result would have been attained had spine characters received special 
weighting when coded. 



284 A SURVEY OF THE TROPICAL GENERA 

Long experience has convinced taxonomists that characters are of unequal value 
in classification ; to produce a reasonable systematic arrangement by taxonometric 
methods they could well be unequally weighted, but unfortunately such weighting 
cannot usually be made before decision as to the nature of the group. Hence, since 
the main purpose of these methods is to aid decision-making, a priori weighting or 
advance decision would reduce their effectiveness ; characters need to be weighted 
a posteriori according to their distribution in the groups formed on the basis of overall 
similarity. Apparently the best course would be to accept as a basis the main 
groupings produced from an unweighted coding, then to investigate the content of 
these groups and modify them, if need be, according to the concepts of unequal 
taxonomic value of characters which emerge from this study. 

GEOGRAPHICAL DISTRIBUTION AND ECOLOGY 

The situation thus demonstrated in Opionia parallels in some respects that de- 
scribed for the perennial species of Ziziphora {Labiatae) by I. C. Hedge (1961). The 
general habitat of these plants of the Near and Middle East is arid steppe and dry 
alpine slopes ; that of Opionia is generally dry rocky slopes. In Ziziphora the plants 
are small-leaved and have small zygomorphic flowers showing sexual states which 
had not previously been recognized as such, again much as in Opionia. The occur- 
rence within a single Ziziphora population of forms having very diverse hair-covering, 
together with scarcely correlated differences in habit of growth, proportion of leaf, 
length of calyx and shape of calyx-teeth, led Hedge to the conclusion that, despite 
some forty specific names applied to these plants, they form one actively evolving 
genetic unit in which certain phenotypic trends are apparent but are not yet marked 
by discontinuities justifying their specific or infra-specific designation. He found it 
impossible to construct a satisfactory key to the variants of Ziziphora but accepted a 
number of reference points within the multiplicity of trends in variation. This 
likewise is largely true of Opionia. Many combinations or intergradations of char- 
acter are possible but some are represented by few or no individuals. Ziziphora 
inhabits a continuous landmass extending from Bulgaria over Turkey, the Caucasus 
and northern Persia to the Western Himalaya and the Altai mountains. The main 
territory of Opionia is the Antillean chain of islands, most of which have long been 
separated from one another and others, e.g. the Lesser Antilles, have always been 
independent (cf. Alston, 1952 ; Schuchert, 1935). Thus conditions here favour 
differentiation. Cuban species, such as Opionia nannophylla and 0. tetrasticha, and 
Jamaican species, such as 0. acicularis, 0. armata and 0. jamaicensis, individually 
possess morphological features unknown in the other islands. There are, however, 
forms extending from the Bahamas to the Virgin Islands and from Cuba to the 
Grenadines which do not make a homogeneous series or even a cline and do not 
break up into consistently distinct populations, but which include here and there 
forms so different that considered apart from the whole assemblage they would seem 
to merit specific rank and have indeed sometimes been given it. In a genus of closely 
allied plants, independent homologous and parallel variation is to be expected ; thus 
similar-looking plants of different immediate phylogeny may occur in areas remote 
from one another. The classification adopted here is, it is hoped, less arbitrary than 



r 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 285 



Lindau's but is likely to be arbitrary nevertheless in that, by attempting to assemble 
specimens linked by overall resemblance into groups which can be recognized by a 
'few fairly well-marked and constant features and be named according to the nomen- 
clatural types coming within their circumscription, no account is or can be taken of 
parallel or convergent evolution ; moreover not all specimens manifest clearly the 
chosen diagnostic features. With these reservations the genus, as stated above, has 
been divided into 14 species, which although not always easy to distinguish seem to 
have geographical as well as morphological justification. 

The genus Opionia as defined and classified here consists of one species {0. grandi- 
flora) in northern Peru, three species (0. acicularis, 0. armata and 0. jamaicensis) 
confined to Jamaica, three species {0. nannophylla, 0. purpurascens and 0. tetrasticha) 
confined to Cuba, two species of much wider Antillean range, i.e. 0. spinosa, which 
extends from the Bahamas over Cuba, Hispaniola and Puerto Rico (but not Jamaica) 
to the Virgin Islands, and 0. microphylla, which extends from eastern Jamaica to the 
Grenadines in the Lesser Antilles, and, by the inclusion of Forsythiopsis, also five 
species from Madagascar. They usually inhabit the drier regions of these islands, 
their normal habitats being on limestone cliffs or rocky slopes thinly covered 
with scrub forest and thus well-drained, open and sunny. The xerophytic features 
characteristic of most species, i.e. the marked development of spines and reduction 
of leaf size and inflorescence, accord with their ecology but are uncommon within the 
family Acanthaceae as a whole. It is essentially these features of reduction which 
separate them from the small-flowered species of Odontonema, such as 0. nitidum, 
which are of mesophytic habit, with relatively large leaves, numerous flowers in 
racemes or narrow panicles and no spines. Psilanthele eggersii of Ecuador approaches 
closest to Odontonema. Within Opionia itself, 0. grandiflora of Peru, 0. jamaicensis 
of Jamaica and 0. acuminata of Madagascar approximate most to the presumed 
archetype of the group in having a mesophytic habit, relatively large leaves, no spines 
and several or many flowers in axillary fascicles which are evidently condensed racemes. 

The occurrence of species showing relatively primitive features, as Psilanthele 
eggersii and Opionia grandiflora, within the region of northwestern South America 
(northern Peru, Ecuador and Colombia ; Text-fig. 9), which has been an area of 
major development for the Acanthaceae — thus 39 native genera of Acanthaceae, with 
365 species, are recorded from Colombia by Leonard (1951-58) — points to this as 
being the area where Opionia had its origin, even though it has reached its greatest 
morphological diversity in Cuba and Jamaica. It misses the Central American main- 
land entirely. In this it is far from unique phytogeographically. As Seifriz (1943) 
mentions, the plant life of Cuba shows great similarity to that of South America, less 
to Central America and little to North America, and the same is probably true of the 
other Greater Antillean islands, Jamaica, Hispaniola and Puerto Rico. There are 
certainly floristic links between South America and the Greater Antilles which are 
unlikely to have been made by way of a former land connection between Honduras 
and the Greater Antilles (regarding which cf. Schuchert, 1935 ; Asprey & Robbins, 
1953)- Thus the genus Neriacanthus [Acanthaceae) consists of N. grandiflorus 
Leonard in Colombia, A^. lehmannianus (Lindau) Lindau in Colombia and Peru, and 
N. purdieanus Benth. in Jamaica. Heppiella (Gesneriaceae) has about ten species in 



286 



A SURVEY OF THE TROPICAL GENERA 



tropical South America, none in Central America, H. corymbosa (Swartz) Urban in 
south-eastern Jamaica and H. cubensis Morton in Cuba (Oriente province) . Huertea 
{Staphyleaceae) has one species, H. glandulosa Ruiz & Pavon, in Peru, none in Central 
America, and one, H. cubensis Urban, in Cuba and Hispaniola. Of the genus 
Purdiaea [Cyrillaceae), according to J. L. Thomas (i960, 1961), there is one species, 
P. nutans Planchon, in Peru, Colombia and Venezuela and another one, P. belizensis 
(Smith & Standi.) J. L. Thomas, in British Honduras, whereas Cuba possesses ten 
species, of which one, P. cubensis (A. Rich.) Urban, grows in the Isle of Pines and 
Pinar del Rio province and of which the other nine are endemic to Oriente province. 




Fig. 9. Distribution of Oplonia and Psilanthele in tropical America ; i, West Indian species 
of Oplonia ; 2, Oplonia grandiflora in Peru ; 3, Psilanthele eggersii in Ecuador. [Goode Base 

Map, University of Chicago) . 



Thomas (i960) states : " there are several trends of specialization in Purdiaea which 
indicate that the South American species, P. nutans, is the most primitive species in 
the genus ". The Melastomataceae provide further examples of fioristic linkage. 
Thus Clidemia grisebachii Cogn. of Jamaica and C. cruegerana of Trinidad and 
Venezuela are closely related, likewise Miconia nubicola Proctor of Jamaica and 
certain species of the Venezuelan coastal Cordillera and the Andes with no close 
relatives elsewhere (J. J. Wurdack in litt.). Henrietta sessilifolia Triana occurs in 
Jamaica, Trinidad and Venezuela, H. ramiflora (Swartz) DC. in Jamaica, Trinidad, 
Venezuela, the Guianas and Brazil. 

Such facts taken from a diversity of families speak for a former direct way of 
fioristic migration between north-western South America and the Greater Antilles in 
which Central America had little or no part. Despite the limited collecting then 
available, Grisebach as long ago as 1865 alluded to such a " besondere Wanderungs- 



OPLONIA AND PSILANTHELE (ACANTH ACE AE) 



287 



linie " connecting the Andes of South America with the mountains of Jamaica and 
Cuba. Croizat (1952 : fig. 71), arguing largely from the distribution of some Cacta- 
ceae, also postulates a Peru-West Indies track across Venezuela direct to the Greater 
Antilles. 

Moreover it seems improbable that the rich and diversified endemic flora of the 
Greater Antilles evolved on so limited an area as this flora now occupies and likewise 
improbable that the ancient floristic connections of the Greater Antilles and northern 
South America were made by long-distance leaping dispersal across the great expanse 
of the Caribbean Sea, although floating seeds carried by sea currents must have taken 
a number of coastal plants from island to island and light air-borne seeds, e.g. of 







A nAAT/C OCfAA/ 




UADCLOUPt ^ 



BARBADOS 



Fig. 10. Distribution of Oplonia nannophylla A ; O. spinosa •. 

members of the Apocynaceae, have been swept over the area by high winds. In 
general, the biological facts seem to favour the hypothesis that " until the end of the 
Cretaceous a continental landmass occupied the position of the present-day Caribbean 
Sea " (cf. Weyl, 1965), and this landmass [Caribia) subsided during the Cretaceous 
and Tertiary periods. The submergence of almost the whole area, including Jamaica 
and Cuba, by Oligocene times would have led to the extinction of the major part of 
its flora but much of the present Greater Antillean flora may nevertheless be descended 
from the plants of this lost territory which presumably both contributed to and 
received from the flora of South America and contributed to that of Central America. 
Such a hypothesis provides a plausible background to facts of Caribbean phyto- 
geography at present otherwise inexplicable, among them the distribution of Opionia. 
Croizat (1952 : 368) has indeed gone so far as to describe " the tale of dispersal in 
the Caribbeans " as " the record how a land once filling the extant sea crumbled 
front by front, thereby creating huge gaps in distribution. " 



288 



A SURVEY OF THE TROPICAL GENERA 



Jamaica and Cuba are the major islands of development for Oplonia. The species 
show different patterns of distribution (see Text-figs. 10-14). 

In Jamaica, 0. microphylla (Text-fig. 12) inhabits the very hot and dry honeycomb 
limestone along the southern coast from St. Thomas to St. Catherine, which has a 




I (ARUMa o 



LAiiqvitt* 

MAROARITA 



Fig. II. Distribution of Oplonia purpurascens ♦ ; O. microphylla 




Fig. 12. Distribution of Oplonia acicularis O ; O. jamaicensis • ; O. microphylla (in 

Jamaica) ♦. 



characteristic scrub forest flora described by Asprey and Robbins (1953)- 0. acicularis 
(Text-fig. 12), with an evident preference for moister and even somewhat shaded 
habitats, has a remarkably disjunct distribution, one area in the east and another in 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 



289 



the west but not in the middle of the island. 0. jamaicensis (Text-fig. 12), considered 
the most primitive Antillean species, grows only in the moist interior upland region, 
0. armaia (Text-fig. 13), as accepted here, presents much diversity in habit and leaf- 
form with geographical differentiation mostly in colour of the corolla. Red flowers 
occur only on plants inhabiting the interior (var. armata), paler bluish or lilac flowers 
only on those nearer or at the coast (var. pallidior) . 




Fig. 13. Distribution of Oplonia armata ; var. pallidior • ; var. armata O. 

I 1 




Fig. 14. Distribution of Oplonia tetrasticha ; var. tetrasticha • ; var. polyece ♦. 

Cuba divides conveniently into four botanical regions (cf. Conde, 1952, also Alain, 
1953. 1958) ; I : Occidente, comprising the provinces of Pinar del Rio, Habana and 
Matanzas ; 2 : Las Villas, the former province of Santa Clara ; 3 : Camaguey ; 
4 : Oriente. 0. purpurascens (Text-fig. 11) is apparently endemic to the mogotes, the 
precipitous limestone " haystack hills" (cf. Seifriz, 1943 : 394 ; Lehmann, 1954) of 
Pinar del Rio around Vinales. 0. nannophylla (Text-fig. 10) occurs in Pinar del Rio, 
Habana and Las Villas. 0. tetrasticha var. tetrasticha (Text-fig. 14), with simple 
spines, likewise occurs in Pinar del Rio, Habana, Las Villas and Oriente, but is largely 



290 A SURVEY OF THE TROPICAL GENERA 

replaced in Oriente province, rightly described by Croizat as a "hotbed of endemism " 
by 0. tetrasticha var. polyece with branched spines. Oplonia has not yet been 
collected in Camaguey. In Oriente occur also plants referred to 0. spinosa (Text-fig. 
lo) (see p. 307), a species which extends over the low-lying Bahama Islands northward 
to Grand Bahama and New Providence and eastward to the Virgin Islands. Within 
Oriente province it appears to be confined to the Moa region on the north coast. 

0. microphylla (Text-fig. 11), which has not colonized the Bahamas, extends the 
range of the genus beyond the Greater Antilles to the Grenadines in the Lesser 
Antilles, which are of volcanic origin, but it does not reach Barbados, Tobago or 
Trinidad. The absence of the genus from these three islands, which lie on the 
continental shelf of north-western South America, as also from Venezuela, makes it 
as improbable that it migrated from South America by way of the Lesser Antilles as 
that it did so along the isthmian route leading to the presumed Antillean-Honduran 
land bridge. 

From the above survey it would appear that the genus Oplonia may have had 
its origin in South America from mesophytic shrubs resembling Psilanthele eggersii 
or small-flowered species of Odontonema, that it reached the Greater Antilles by way 
neither of Central America and the Honduran land-bridge nor of the Lesser Antilles 
but by a former direct long- vanished link, and that its major development has taken 
place in Jamaica and Cuba with little change in floral structure but with increasing 
vegetative adaptation to xerophytic conditions. 

PHYTOGEOGRAPHICAL LINKAGE OF TROPICAL AMERICA AND MADAGASCAR 

A remarkable discontinuity of range associated with little morphological diver- 
gence becomes apparent when the American species of Oplonia are compared with 
the Madagascan species currently referred to Forsythiopsis. Since their resem- 
blances in habit, floral structure, capsule and pollen together leave no characters 
available for generic separation, they are here placed in the same genus despite their 
geographical remoteness. Use of different generic names for the American and 
Madagascan plants respectively has obscured their affinity. Baker would surely 
never have described Forsythiopsis as an independent genus in 1883 had he not then 
compared Baron's Madagascan specimens, the type material of Forsythiopsis baronii 
Baker, with Ruellia instead of with Anthacanthus (i.e. Oplonia), to which Nees von 
Esenbeck had already in 1847 referred Lamarck's conspecific Justicia vincoides, 
likewise from Madagascar. Moreover, the pollen of Forsythiopsis, as is evident from 
that of the holotype oiF. baronii {Baron 1937), the holotype of F. linifolia {Humbert & 
Swingle 55oy), an authentic specimen of F. australis {Scott-Elliott 2jo) and other 
specimens {McWhirter & Capuron 184, Humbert 2606) studied with the scanning 
electron microscope at magnifications between x 500 and 5,400, agrees with the 
pollen of West Indian Oplonia specimens ; it is prolate or spheroid, 3-colporate, 
each colpus flanked by a pseudo-colpus extending almost to the poles, the surface 
pitted with minute lumina (PI. 41) ; it in no way resembles the reticulate pollen of 
Ruellia. Similar pollen occurs in Pseuderanthemum and also the continental African 
genus Ruttya, within which C. B. Clarke would have incorporated Forsythiopsis, 
but this has larger flowers with monothecous (unilocular) anthers (whence Hoch- 



I 



OPLONIA AND PSILANTHELE (ACANTH ACEAE) 



291 



stetter's name Hapianthera), possibly derived from a P sender anthemum- or a For- 
sythiopsis-like stock with dithecous (bilocular) anthers. 



I 



12 


n '^^° 


46 48 


50° 




^ 










12 


c <J 








c/ 


'^ 















J 


s 


\ 




14 








( 




A 


\ 


14 








( 




15 








0' 




c\ 


\ 


15° 








L/^ 


) 


16 




/\t 


■f^ 


H 


\V 


16 


r^ 




• 


"i 




) 


• 


k 








/ 


• • 


\ 






18 


] 




) 




18 




^ • 


• 

Tananarive 

/ 


1 






?n° 


\ 


/ 






?n° 


?? 


/ ■ 1 




9? 




\* +♦ 


*" 


r 










V 


1 










24 
2^ 


1 


> 


• 
♦ 


1 








24 
25° 


\t' 




phin 




H^^ 


SJST^ 






Stalule MHes 














20 40 60 80 100 




45° 


46 48 


50° 





Fig. 15. Distribution of Oplonia in Madagascar ; Oplonia vincoides • ; O. minor var. 
minor ♦ ; O. minor var. meridionalis o ; O. minor var. vestita * ; O. linifolia ■ ; 
O. acuminata + ; O. puherula A. 

In Madagascar (Text-fig. 15) the species grow mostly within Humbert's phyto- 
geographical domains Sud and Quest (cf. Blanc, 1971, for maps and climatological 
data), arid and semi-arid regions with open xerophytic scrub and savanna vegetation; 



292 A SURVEY OF THE TROPICAL GENERA 

they thus appear to have much the same ecological requirements as most West Indian 
species of Oplonia. Madagascar, with some 400 species of Acanthaceae to an area of 
228,000 square miles, has been a major centre of development for the family, out- 
rivalling even Colombia, with some 370 species to 462,000 square miles. 

No species congeneric with Oplonia (including Forsythiopsis) are known from 
continental Africa. In manifesting such discontinuity Oplonia does not stand 
alone. There are other links, few but definite, between the floras of tropical 
America and Madagascar which testify to a much wider primaeval range of certain 
groups forming part of Perrier de la Bathie's " element austral " in the Madagascar 
flora, " temoins residuels d'une fiore antique en voie d'extinction " (Perrier de la 
Bathie, 1936 : 139-140). Thus Ravenala {Musaceae) has one species, R. madagas- 
cariensis Sonnerat, the well-known " Traveller's Tree " in Madagascar ; its only 
close ally is a species of tropical America (Brazil, Guiana), R. guianensis (L. C. Rich.) 
Petersen ; the latter having five stamens (instead of six) can be put in a genus by 
itself as Phenakospermum guianense (L. C. Rich.) Miq., but, whether the two are 
treated as forming one genus with two species or as two monotypic genera, their 
affinity and their distinctness from all other genera remain beyond dispute. Oliganthes 
{Compositae), with 9 species in Madagascar and 12 in tropical America but none 
elsewhere, provides another example of such linkage, as do Clara {Herreria Ruiz & 
Pavon, non Adans. ; Liliaceae) with 5 or 6 species in South America and Herreriopsis 
with one species in Madagascar. Gladiopappus {Compositae) of Madagascar is 
allied exclusively to the tropical American genera Richterago {Seris Less, non 
Willd.), Hyalis and Hecastocleis. These anomalous ranges only become explicable 
when considered in relation to those of genera or groups of closely allied genera 
occurring not only in tropical America and Madagascar but also in continental 
Africa. For example, the genus Sabicea (including Stipularia; Rubiaceae) has about 
50 species in tropical America (including Peru, Ecuador, Colombia and the West 
Indies), 85 in continental tropical Africa and 5 in Madagascar. A year before the 
pubUcation of Alfred Wegener's book on continental drift, and long before it had 
become generally known, characters common to various American and African 
species of Sabicea led Wernham (1914) to the opinion that " the genus existed simul- 
taneously in Africa and America as an ancestral race [Primosabicea] with lax in- 
florescence, represented at the present day by e.g. 5. venosa in Africa and 5. paraensis 
in America ".It has maintained a continuity of range in Africa, unlike, for example, 
the genus Hirtella [Chrysobalanaceae) with approaching 100 species in tropical 
America, 2 in tropical East Africa, of which one occurs in Madagascar but none in 
West Africa. Mendoncia {Acanthaceae) has about 90 species in tropical America, 
4 in tropical West Africa (but not East Africa) and 3 in Madagascar. Symphonia 
{Guttiferae) is remarkable in having one species (5. globulifera L.f.) common to 
tropical America and tropical West and Central Africa and 15 species recorded from 
Madagascar. An even more remarkable range has been attained and maintained 
by Christiana africana DC. {Tiliaceae ; cf. Kew Bull. 17 : 501 (1964)), which occurs 
in tropical America, tropical West, Central and East Africa and Madagascar ; 
the one other species of the genus is Brazilian. Genlisea {Lentibulariaceae) , with 
15 species, also occurs in tropical America, tropical continental Africa and Mada- 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 293 

gascar. Savia [Euphorbiaceae) has two main centres of diversity, with 12 species in 
the West Indies and 9 in Madagascar, but their discontinuity is sHghtly lessened by 
2 species in Brazil and one in South Africa ; one species extends into southern 
U.S.A. The Velloziaceae display a likewise significant pattern of distribution 
linking tropical America, continental Africa and Madagascar, as noted, for example, 
by Camp (1947). The family consists of Vellozia and Barhaceniopsis, which are 
exclusively tropical American, and Barhacenia (sensu lato, including Xerophyta) 
which has about 75 species in tropical America, about 40 species in continental 
Africa, 3 in Madagascar, one in southern Arabia. Several consistent features distin- 
guish American Barhacenia (sensu stricto) from African Xerophyta ; one African 
species {Vellozia elegans Oliver of Natal), in the opinion of L. B. Smith and E. Ayensu 
(in lift.), constitutes an intermediate monotypic genus [Talbotia Bali, f., non S. 
Moore) combining several primitive features, which suggest that the family originated 
in Africa ; but whether the African species are included in Barhacenia or separated 
from it generically is irrelevant from a phytogeographical standpoint in view of the 
undoubted affinity of the American and African groups. A somewhat like affinity 
is that of Aphelandra {Acanthaceae) , having about 200 species in tropical America, 
with Stenandriopsis having 5 species in tropical West Africa, one in East Africa and 
8 in Madagascar (cf. Heine in Aubreville, Fl. Gabon 13 : 99 ; 1966). The genera 
Agarista (Ericaceae) and Agauria provide another example ; Agarista (cf. Stevens, 
1970) has about 34 species, mostly in South America but with a few in Central 
America and the southern United States ; the closely allied genus Agauria, according 
to Sleumer (1938), consists of one polymorphic species divisible into 20 intergrading 
infraspecific taxa, a few occurring widely separated in West Africa, more on the 
East African mountains, most of them in Madagascar. 

Equally significant is the existence of other groups connecting tropical America 
and Madagascar which have more restricted ranges evidently resulting from the 
fragmentation and reduction of former more extensive ranges. Thus the genus 
Erhlichia (or Piriqueta sect. Erhlichia ; Turner aceae) has one species in Central 
America, one in southern Africa, three in Madagascar. Another phytogeographic- 
ally relevant group is that formed of three closely allied monotypic genera of 
Compositae described by Lawralree (1943), i.e. Eleutherantha in tropical America, 
Hoffmanniella in Cameroons and Congo, Exomiocarpon in Madagascar. A rather 
similar linkage is to be found in the Dichapetalaceae with Stephanopodium in tropical 
America, Tapura in tropical America and continental Africa, Falya in Madagascar. 
The Canellaceae, one of the nine exclusively Amerafran families*, exemplifies the 
same general floristic linkage, with Canella, Cinnamodendron and Pleodendron in 
tropical America, Warburgia in tropical East Africa and Cinnamosma in Madagascar. 
The genus Carpodiptera {Tiliaceae) does not reach Madagascar but nevertheless 
manifests a distribution of the same type as Hirtella by having species in Central 
America and the West Indies, tropical East Africa, and the Comoro Islands 

* Angiosperm families occurring both in tropical America and Africa but not known elsewhere are 
Bromeliaceae, Canellaceae, Caricaceae, Humiriaceae, Hydnoraceae, Mayacaceae, Rapataceae and Vochysi- 
aceae (cf. Vaster, 1940 : 527-529, 567-569, maps 238-241, 245-248). Loasaceae and Velloziaceae occur in 
tropical America and Africa but extend from east Africa into southern Arabia (cf. Vester, 1940 ; maps 
243. 244). To these can be added Strelitziaceae if separated from Musaceae. 



294 A SURVEY OF THE TROPICAL GENERA 

north of Madagascar (cf. map in J. Hutchinson, Evolution and Phytogeny of 
Flowering Plants : 232 (1969)). 

There are also families and genera occurring both in tropical America and West 
Africa (cf. Hepper, 1965) which do not reach Madagascar but nevertheless emphasize 
the floristic affinity of the two continents. One such is Cienfuegosia (cf. Fryxell, 
1969) in the Malvaceae with 16 American and 8 African species. 

Taken together these diverse ranges form a series manifesting stage by stage 
the reduction from a wide continuous range, such as that of Sabicea (America, conti- 
nental Africa, Madagascar), by way of intermediate discontinuous ranges, such as 
those of Erblichia and Hirtella, to the extremely discontinuous ranges of Ravenala, 
Oliganthes and Opionia. These floristic linkages must have had their origin in the 
far distant past, presumably long before the end of the Cretaceous period, when the 
continuity of the Amerafran landmass (cf. maps in Bullard, Everett & Smith, 

1965 ; Smith & Hallam, 1970 ; Dietz & Holden, 1970), later separated into South 
America and Africa, still provided the opportunity for (or the narrowness of the 
South Atlantic Ocean made no obstacle against) the spread of extreme western- 
originated groups eastward across that vast landmass, the interior of which must 
have been very dry, however, before separation, and the spread of eastern-originated 
groups westward, since climatic conditions were presumably more favourable for the 
evolution and preservation of diversity towards the periphery than towards the 
centre. If so, the groups concerned have kept their generic characters, or at least many 
characters indicative of affinity, essentially unchanged for probably 90 million years 
or more ; indeed Hawkes and Smith (1965) have argued from the taxonomy and 
distribution of Bromus, Gossypium and Solanum that these genera existed already in 
the early Cretaceous, some 100 to possibly 135 milHon years ago. 

Madagascar has certainly been isolated for a long time. If, as stated by Flories 
(1970), the rift which became the Mozambique Channel separating Madagascar from 
continental Africa had been completed by the end of the Cretaceous period, i.e. some 
70 million years ago, at which time modern orders and families of angiosperms were 
already differentiated (cf. MuUer, 1970), then the ancestors of the Madagascan plants 
whose nearest allies now inhabit tropical America must have already acquired the 
characters still common to the Madagascan and American plants or at least a close 
genetical similarity determining later parallel development. This conclusion, if correct, 
implies the origin of the Acanthaceae long before that part of the Miocene epoch, some 
12-20 milhon years ago, to which belongs the earliest available palaeontological evi- 
dence of its existence, i.e. pollen of iho. Justicia-iy^e named Multiareolites formosus 
and of the Trichanthera-type named Multimarginites vanderhammenii (cf . Germeraad 
et al., 1968: 302, t.6) from Venezuela; phytogeographical evidence indicates a 
Cretaceous origin. 

Acceptance of the view that the angiosperms originated on the Gondwanaland 
continent and that many genera existing today had acquired their distinctive 
characteristics before it broke into several continental landmasses (cf. Melville 

1966 ; Schopf, 1970) harmonizes many otherwise seemingly inexplicable facts of 
angiosperm distribution. The drifting apart of Africa and South America, together 
with the associated far-reaching climatic disturbance, must have led to the extinction 



i 



OPLONIA AND PSILANTHELE ( ACANTHACEAE) 295 

or reduction of range of some groups while promoting the development and spread 
of others. Evidently those groups such as Oliganthes and Oplonia, now with 
representatives in tropical America and Madagascar but not elsewhere, owe the 
isolation of these to continental drift which divided the original stock into American 
and African populations, to the spread of the African population across Africa and 
over to Madagascar or of a Madagascar-originated population across Africa to America, 
to climatic changes, which led to extinction in continental Africa, and to peripheral 
survival in America and Madagascar, with the larger area and more varied habitats 
of tropical America favouring greater specific development and range extension. 

ACKNOWLEDGMENTS 

The preparation of this survey of Opionia and Psilanthele would probably have been 
simpler had a revision of the whole group been planned from the start instead of 
undertaken reluctantly as the only way of elucidating nomenclatura] problems 
affecting Jamaican species. No attempt has been made to borrow for study all the 
material available from islands other than Jamaica. My thanks are due to the 
directors and curators of the following institutions for the privilege of examining 
specimens in their charge : the Arnold Arboretum of Harvard University, Cambridge, 
Mass. (cited as A) : the Field Museum of Natural History, Chicago (F) ; the Gray Her- 
barium of Harvard University (GH) ; the Science Museum, Institute of Jamaica, 
Kingston, Jamaica (I J) ; the Royal Botanic Gardens, Kew (K) ; Herbario de la Salle, 
Habana (LS ; temporarily deposited in NY) ; the Institute Botanique, Universite 
de Montreal, Montreal (MT) ; the New York Botanical Garden, New York (NY) ; 
Laboratoire de Phanerogamic, Museum National d'Histoire Naturelle, Paris (P) ; 
the Botanical Department, Naturhistoriska Riksmuseum, Stockholm (S) ; Depart- 
ment of Botany, University of the West Indies, Mona, Jamaica (UCWI). 

I am also grateful to Brother Alain (H. Liogier), formerly of the Colegio de la Salle, 
Habana, for details of distribution in Cuba, to the Director, Botanisches Museum, 
Berlin-Dahlem, for photographs and to Dr. G. Wagenitz for critical details of the 
types of Eranthemum emarginatum and E. lycioides in the Willdenow Herbarium, to 
Professor G. Erdtman, Stockholm-Solna, for preparations of pollen grains, to the late 
Mr. J. B. Evans, British Museum (Natural History), also for preparations of pollen 
grains and of cleared leaves, to Dr. Jose de Js. Jimenez, Santiago de los Caballeros, 
for specimens from the Dominican Republic, and to Mr. G. R. Proctor, Institute of 
Jamaica, for companionship and help on many a field trip in Jamaica and for much 
information about its floristics and geography. Dr. H. Heine and Prof. J. Leandri 
gave valuable help in mapping the distribution of Opionia in Madagascar. 

The taxonometric section of this paper owes its existence to the co-operation of 
Dr. P. H. A. Sneath, British Medical Research Council, Mr. J. G. Gower and Mr. 
G. J. S. Ross of the Rothamsted Experimental Station, Harpenden, Herts., England 
and Professor D. J. Rogers, University of Colorado, Boulder, Colorado, U.S.A. ; 
their computation and analysis of the data supplied and preparation of diagrams 
have proved stimulating and informative and this acknowledgement is only a mild 
expression of gratitude for their help. 

The photographs of pollen grains illustrating this paper have been made with the 



296 



A SURVEY OF THE TROPICAL GENERA 



" Stereoscan " electron microscope in the Electron Microscopy Unit, British Museum 
(Natural History) and I thank Miss P. A. Sims and Dr. S. Sengupta for their valuable 
technical assistance in this and other palynological work. 

Miss Beatrice M. Corfe drew Text-fig. i, Miss Joanna C. Webb (Mrs. D. Erasmus) 
Text-figs. 2 and 3, Mr. Derrick Erasmus Text-figs. 4 and 16 and Miss Victoria 
Goaman Text-figs. 17 and 18. 

PSILANTHELE Lindau 

PsiLANTHELE Lindau in Bull. Herb. Boiss. 5 : 663 (1897). 

Shrubs without spines. Leaves small, shortly petiolate, entire ; cystoliths 
present. Flowers small, pedicellate in short-pedunculate few-flowered loose 
raceme-like cymose inflorescences terminating short axillary branches. Bracts 
minute. Calyx 5-parted, c.2 mm. long, the tube very short, the segments almost 




Fig. 16. Flower of Psilanthele eggersii Lindau {Eggers I5i2g) cut lengthwise. 

equal, narrow, acute. Corolla zygomorphic, bluish, c.5 mm. long ; tube narrow,! 
cylindric, almost straight ; limb spreading, 2-lipped, the upper lip shortly 2-lobed,| 
the lower lip 3-parted, the segments almost equal. Stamens 2, inserted above the] 
middle of the corolla-tube ; filaments short ; anthers with 2 spurless equal disjunct] 
loculi ; staminodes 2, extremely minute. Pollen prolate, 3-colporate, each colpusj 
flanked by 2 pseudocolpi, reticulate with minute lumina. Ovary with 2 ovules in] 
each loculus ; style relatively long. Capsule clavate, stipitate, 4-seeded. Type- 
species by monotypy : Psilanthele eggersii Lindau. 

Only known species the following : 
I. Psilanthele cggemt Lindau in Bull. Herb. Boiss. 5 : 664 (1897). — Mildbraed apud] 
Diels in Bibl. Bot. 116 : 148 (1937). (Text-fig. 16 ; Pis. 40a, 46). 

Shrub to 2 m. high ; branches slender, ascending, sparsely and minutely pubescent,] 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 297 

red-brown (fide Diels) when young, becoming yellowish with age. Leaves short- 
stalked, the blade narrowly elliptic or lanceolate, the apex acuminate, the base 
attenuate, c. 1-5-5 cm. long, 5-15 mm. broad, glabrous and with numerous cystoliths 
above, glabrous below, thin, the veining evident. Flowers in 16-20-flowered 
inflorescences 1-2 cm. long ; pedicels 2-3 cm. long. Calyx c.2 mm. long, the seg- 
ments very narrowly triangular. Corolla blue (fide Lindau) or white with red 
streaks on lower lip (fide Diels), minutely pubescent, c.5 mm. long, the tube 2-5 mm. 
long, the segments 2-5 mm. long. Stamens almost included, the filaments 07 mm. 
long, the anthers i mm. long. 

Ecuador : Manabi : probe Hacienda El Recreo, 16 July 1898, Eggers I5i2g 
(F ; K ; isotypes). Chimborazo : im Gebiet des R. Chanchan bei Naranjapata, 
530m., Sept. 1933, Diels iig2 (fide Mildbraed, loc. cit). 

The above account is based on Lindau's original description and isotypes in the 
Kew and Chicago herbaria. As the type was collected by Baron Henrik Franz 
Alexander Eggers (1844-1903) in western Ecuador near the coast at El Recreo 
(0° 27'S) south of Cape Pasado, and Mildbraed records it from central Ecuador at 
Naranjapata (2° 17'S), it would seem to be a species widely distributed in Ecuador 
but rarely collected, the flowers being inconspicuous and its habit of growth in no 
way remarkable. Eggers was a Danish professional soldier, who served in Mexico 
and the Danish West Indian islands (St. Croix, 1869-72, 1873-74, St. Thomas, 
1872-73, 1874-1885) and who managed while so engaged to collect botanical speci- 
mens on many West Indian islands. In 1891-99 he travelled and collected in South 
America (Venezuela, Ecuador). An account of his career will be found in Urban, 
Symh. Antill. 3 : 40-43 (1902). 

A photograph of the holotype (formerly in Berlin-Dahlem) is available in Field 
Museum of Natural History, Chicago, photograph 8737. 

OPLONIA Raf. 
Oplonia Raf., Fl. TeUur. 4 : 64 (1838). 

Eranthemum c. Hesperanthemum Endl., Gen. PI. : 706 (1839). 

Anthacanthus Nees in DC, Prodr. 11 : 460 (1847) nom. superfl., excl. § 2. — Griseb., Fl. Brit. 

W. Ind. Is. : 457 (1862). — Benth. in Benth. & Hook., Gen. PL 2 : 1097 (1876). — Lindau 

in Engler & Prantl, Nat. Pflanzenfam. iv. 3b : 335 (1895) ; in Urban, Symb. Antill. 

2 : 214 (1900). — Britton & Millsp., Bahama Fl. : 402 (1920). — Britton & Wils. in Sci. 

Surv. Porto Rico & Virgin Is. 6 : 212 (1925). — Alain in Le6n & Alain, Fl. Cuba 4 : 498 

(1957)- 
Forsythiopsis Baker in Joum. Linn. Soc. Lond., Bot. 20 : 218 (1883). — Benoist in Humbert, 

Fl. Madag. 182 : 106 (1967). 
Hesperanthemum (Endl.) Kuntze, Revis. Gen. PI. 2 : 490 (1891), nom. superfl. 

Shrubs or small trees, mostly with opposite axillary spines, sometimes spineless. 
Leaves usually small, shortly petiolate or almost sessile, entire, often fascicled when 
two or more pairs arise on the same extremely short axillary branch ; cystoliths 
present, often abundant. Flowers small, pedicellate, solitary and axillary, or in 
sessile or short-pedunculate axillary fascicles. Bracts minute. Calyx 5-parted, 
3-6 mm. long, the tube very short, the segments almost equal, narrow, acute. 



298 A SURVEY OF THE TROPICAL GENERA 

Corolla zygomorphic, white, bluish, purple or red, 8-18 mm. long ; tube narrow, 
cylindric, almost straight, slightly expanded above ; limb spreading, 2-lipped, the 
upper lip 2-lobed, the lower lip 3-parted, the segments almost equal ; aestivation 
imbricate. Stamens 2, inserted above the middle of the corolla-tube, included or 
exserted ; filaments slender, short or long ; anthers with 2 parallel spurless equal 
loculi ; staminodes 2, minute. Pollen spheroid or prolate, 3-colporate, each colpus 
flanked by 2 pseudocolpi, pitted with minute lumina. Ovary with 2 ovules in 
each loculus ; style long, the stigma slightly thickened and almost entire. Capsule 
clavate, stipitate, 2-4-seeded. Seeds somewhat flattened, almost circular in outline, 
2-3 mm. broad, borne on acute retinacula. 

Type-species by original designation : Oplonia spinosa (Jacq.) Raf. based on 
Justicia spinosa Jacq. 

Species variable and difficult to define, fourteen accepted here, one from Peru, eight 
West Indian, ranging from the Bahamas and Cuba to the Grenadines in the Lesser 
Antilles, and five Madagascan. 

As Brizicky (1969) has pointed out, it is impossible to decide the rank of Endlicher's 
subdivisions of genera. 

Key to Species of Oplonia 

Spines up to 3 cm. long and mostly more than 1-5 cm. long, spreading horizontally or 

slightly ascending, straight, rigid ; Cuba . . . 6. 0. tetrasticha 

Spines all unbranched ....... 6a. var. tetrasticha 

Spines mostly bearing 2-4 subsidiary spines . . . 6b. var. polyece 

Spines absent, or, if present, mostly less than 1-2 cm. long, or, if longer, deflexed or 
curved or very slender : 
Spines present, some or all slightly or distinctly curved, spreading almost horizont- 
ally or recurving ; Bahamas to Virgin Islands . . 4. 0. spinosa 
Spines absent or, if present, straight and spreading horizontally or ascending : 
Shrub of very dwarf habit, 4-8 cm. high, with internodes 1-3 mm. long; Cuba. 

9. 0. nannophylla 
Shrub taller and looser in habit, the internodes more than i cm. long : 
Pedicels shorter than the calyx : 

Spines present ; western Cuba . . . . 8. 0. purpurascens 

Spines absent ; Madagascar . ..... i^.O. minor 

Pedicels some or all longer than the calyx : 

Corolla white ; stamens always included, the filaments c.o-8 mm. long, 
shorter than the anthers ; low spreading shrub under 60 cm. high, with 
leaves mostly broadest below the middle ; Jamaica 7. 0. acicularis 
Corolla red, purple, pale violet or blue, very rarely white ; stamens exserted 
or included, the filaments 1-5-7 mm. long, longer than the anthers ; 
shrubs mostly 1-2 m. or even small trees to 6 m. high, with leaves 
variable but often broadest at or above the middle : 
Leaves all very small (not more than 12 mm. long, 6 mm. broad): 
Leaves linear; Madagascar . . . . . 14. 0. linifolia 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 299 

Leaves elliptic to narrowly obovate : 

Spines present ; Jamaica to the Grenadines . 5. 0. microphylla 

Spines absent ; Madagascar . . . . 12. 0. minor 

Leaves larger, some or all more than 12 mm. long or, if less, then broadest 
at or below the middle: 
Apex of leaf acuminate or acute ; calyx segments about as long as 
the 4-5-5-mm. long corolla tube: 
Flowers 7-10 at each axil ; leaf blade narrowly cuneate at base ; 
Peru . . . . . . . 1. 0. grandiflora 

Flowers 1-3 at each axil ; leaf blade rounded at base ; 

Madagascar . . . . . 10. 0. acuminata 

Apex of leaf usually obtuse or rounded, rarely acute ; calyx segments 
shorter than the 5-10-mm. long corolla tube: 
Corolla segments longer than tube ; style hairy ; Madagascar : 
Mature leaves mostly 3-9 cm. long: 

Calyx segments narrowly triangular, sparsely hairy. 

11. 0. vincoides 
Calyx segments narrowly oblong, densely hairy. 

12. 0. puberula 
Mature leaves less than 3 cm. long . . . 13. 0. minor 

Corolla segments slightly shorter than tube ; style glabrous; 
Jamaica : 
Leaves mostly 3-10 cm. long, i •8-4-5 cm. broad ; flowers in each 
fascicle numerous (4-15) ; interior of Jamaica, mostly 
above 500 m. . . . . . 2.0. jamaicensis 

Leaves smaller, mostly 1-2-5 cm. long, 0-4-2 cm. broad, if longer 

then broadest above the middle and flowers solitary ; 

flowers solitary or few (2-4) in each fascicle. 3. 0. armata 

Corolla red ; leaves often broadest above the middle; plants 

of the interior . . . . .3a. var. armata 

Corolla pale purple, lilac or almost white : leaves mostly 
broadest at the middle ; plants occurring below 500 
m., usually coastal . . .__ 3b. var. pallidior 

I. Oplonia grandiflora (Lindau) Steam, comb. nov. (Pis. 40b, 42). 

Psilanthele grandiflora Lindau in Bull. Herb. Boiss., Ser. 2, 4 : 401 (1904). 

Shrub 1-4 m. high ; branches stout, greyish, at first sparsely pilose with long hairs, 
later glabrous ; spines absent. Leaves short-stalked, the blade very variable in size 
in the same fascicle, mostly lanceolate, sometimes narrowly elliptic, the apex acute, the 
base narrowly cuneate, c.2-10 cm. long, 0-5-3 cm. broad, glabrous except for sparse 
more or less appressed hairs on the veins below and along the margin, the veining 
evident with 6-8 lateral veins each side of the midrib ; petiole 5-10 mm. long. 
Flowers 7-10 in sessile axillary fascicles ; pedicels 1-2 cm. long, glabrous. Calyx 
4-5-6 mm. long, the segments linear, acute. Corolla white with violet markings (fide 



300 A SURVEY OF THE TROPICAL GENERA 

E. Ule), glabrous, the tube 4-4-5 mm. long, the upper lip slightly 2-lobed, c. 5 mm. 
long, the lower lip 3-parted with obtuse segments to 6-5 mm. long, 4 mm. broad. 
Stamens included, the filaments c. i mm. long, the anthers scarcely i mm. long. 
Endemic to Peru. 

Peru : San Martin : prope Fuan-Guerra ad Tarapoto, " Strauch 1-4 m., Bl. weiss 
mit violetter Zeichnung ", Oct. 1902, E. Ule 64g3 (K, isotype). 

0. grandifiora is a geographically isolated species differing from other members of 
the genus in its acute leaves, its linear more or less parallel-sided calyx-segments, 
which equal the corolla-tube in length, and its smaller corollas, but it approaches 
0. jamaicensis in its comparatively large and broad leaves and many-flowered 
fascicles. There seems no adequate basis for the generic separation of this Peruvian 
species from the Jamaican species which Lindau associated with it under the name 
Psilanthele, even though this raises the problem of the presence of ancient South 
American elements in the Jamaican flora (see p. 285). 

When transferred to the genus Opionia, the epithet grandifiora is misleading as the 
species has smaller flowers than any other. Lindau, however, contrasted it not with 
members of Opionia {Anthacanthus) but with Psilanthele eggersii, which has even 
smaller flowers. 

A photograph of the holotype (formerly in Berlin-Dahlem) is available in Field 
Museum of Natural History, Chicago, photograph 8738. 

2. Opionia jamaicensis (Lindau) Stearn, comb. nov. (PI. 39). 

Psilanthele jamaicensis Lindau in Urban, Symb. Antill. 5 : 501 (igo8). 

Sparse straggling shrub to 4 m. high ; branches rather stout, light brown and 
minutely pubescent with very short upcurved hairs when young, later greyish and 
glabrous, somewhat broadened and flattened below the nodes ; spines usually absent. 
Leaves short-stalked, the blade narrowly obovate, oblanceolate or elliptic, the apex 
rounded or obtuse and often emarginate, sometimes mucronate, the base cuneate or 
attenuate, 3-10 cm. long, i -8-4-5 cm. broad, glabrous above, sometimes sparsely 
pubescent along the nerves below, the veining usually prominent below ; petiole 
C.5 mm. long. Flowers 4-15 in sessile axillary fascicles ; pedicels 5-12 mm. long, 
minutely pubescent or glabrous. Calyx c.5 mm. long, the segments very narrowly 
triangular or subulate. Corolla purplish or lavender blue with purple blotches at the 
base of the lobes, pubescent, the tube c.io mm. long, the segments 7-9 mm. long. 
Stamens almost included or long-exserted, the filaments c.2 mm. or 6 mm. long, the 
anthers 1-5 mm. long. Capsule c. 1-5-2 cm. long. 

Endemic to inland woods of Jamaica at about 350-800 m. 

Jamaica : St. James : White Rock Hill, i mUe S. of Sweetwater, 640 m., 10 Mar. 
1956, Stearn 454 (BM ; S), Proctor 11745 {I J). Trelawny : Tyre, near Troy, 600 m., 
24 Apr. 1906, Harris g362 (BM ; K ; NY) ; 3 Apr. 1917, Perkins 1323 (K). 2 miles 
N. of Troy, 530 m., 14 Mar. 1955, Proctor gg4g (IJ) ; \ mile N. of Troy, 480 m., 14 Mar. 
1955, Proctor gg24 (IJ). Manchester : Vicinity of Mandeville, Feb. 1910, 5. Brown 
264 (NY). Clarendon : Peckham Woods, 760 m.. May 1912, Harris 11060 (NY) ; 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 301 

TO Mar. 1954, Proctor 8433 (BM ; I J) ; May 1955, Robertson igg2 (UCWI). Aenon 
Town to McKay, 600 m., Apr. 1962, Adams 11034 (UCWI). St. Ann : Mason River 
District, 3-4 miles NW. of Kellits P.O., 2 Mar. 1957, Proctor 16216 (IJ). Along road to 
HoUymount, Mount Diablo, 820 m., 6 Aug. i%y , Proctor 16523 (I J) ; 680 m.. Mar. 1962, 
Adams 10830 (UCWI). 5^. Catherine : HoUymount, Mount Diablo, 760 m., 11 Aug. 
1908, Harris 8g68 (NY, lectotype ; BM ; F) ; 15 Feb. 1905, Harris 8888 (BM; NY). 
This species was first collected by Swartz but was confused, it would seem, with 
0. armata. It stands apart from other Jamaican species in its large leaves and 
numerous flowers. The corolla is described as " lavender blue " by Perkins, as 
" purple " or " violet " or " pale lavender " by Proctor, as " purple " by Robertson. 

3. Oplonia armata (Swartz) Steam, comb, no v. 

Justicia armata Swartz, Nov. Gen. & Sp. PI. : 13 {1788) ; Fl. Ind. Occ. i : 28 (1797). — ^Vahl, 

Enum. PI. I : 168 (1804). 
Eranthemum armatum (Swartz) R. Br. ex Roem. & Schult. in L., Syst. Veg., ed. nov. i : 176 

(1817). 
Anthacanthus armatus (Swartz) Nees in DC, Prodr. 11 : 460 (1847). — Griseb., Fl. Brit. 

W. Ind. Is. : 457 (1862). — Lindau in Urban, Symb. Antill. 2 : 217 (1900). 
Hesperanthemum armatum (Swartz) Kuntze, Revis. Gen. PL 2 : 490 (1891), excl. syn. Neesii. 



I 



^^ Sparse straggling or erect shrub or small tree (in var, pallidior) ; branches slender, 
arching or spreading, pubescent when young on two opposite sides with short or very 
short ascending or upcurved hairs, the other two opposite sides almost or quite 
glabrous ; spines often absent but when present slender, rigid, 5-10 mm. (rarely to 
20 mm.) long, spreading at an angle of c.75°, glabrous or minutely pubescent. Leaves 
short-stalked, very variable in size and shape, the blade ovate or narrowly ovate, 
elUptic or narrowly obovate or oblanceolate, the apex obtuse or rounded, usually 
emarginate, the base cuneate or attenuate, 0-5-4 cm. long, 0-3-2 cm. broad, glabrous 
above, sometimes sparsely pubescent along the nerves below, the veining below 
scarcely evident to prominent ; petiole 2-5 mm. long. Flowers solitary or 2-4 in 
sessile axillary fascicles ; pedicels 5-15 mm. long, glabrous or minutely pubescent. 
Calyx 4-6 mm. long, the segments very narrowly triangular or subulate. Corolla 
deep pink, rose or red-purple (in var. armata), lavender-blue, pale purple or almost 
white (in var. pallidior), pubescent or glabrous, the tube 8-10 mm. long, the segments 
4-8 mm. long. Stamens almost included or long-exserted, the filaments c. 1-5-2 mm, 
or 5-7 mm. long, the anthers c. 1-1-5 i^^- lo^ig- Capsule c. 1-5-2 cm. long. 

j Endemic to Jamaica, var. pallidior at sea-level to 400 m., var. armata at 150-850 m. 

3a. Oplonia armata var. armata (Text-fig. 3 C, D ; PI. 43b). 

I Shrub 1-5-3 m. high. Leaf-blades narrowly ovate, narrowly elliptic or narrowly 
i obovate, the base cuneate or attenuate. Flowers mostly solitary, sometimes 2-4 in a 

fascicle. Corolla red, variously described as deep rose-pink, red-purple, cerise and 

crimson. 

Forma 1. Leaf-blades vciostlYmiiowlyoYdle, 1-2-5 cm. long, 0-5-1-5 cm. broad. 
Corolla deep pink or bright cerise. 



302 A SURVEY OF THE TROPICAL GENERA 



I 



Jamaica : 5^. James : White Rock Hill, i mile S. of Sweetwater, 640 m., 10 Mar. 
1956, Steam 435 (BM). St. Elizabeth : Balaclava, 150 m., 25 Mar. 1917, Perkins 
1413 (GH ; K). Trelawny : Boothe district, 3 miles N. of Troy, 490 m., 14 Mar. 1955, 
Proctor gg^y (IJ). Troy, 13-18 Sept. 1906, Britton 58g (F ; NY). Manchester : 
Halifax, 26 July 1952, West & Arnold 8g6 (BM). Clarendon : Peckham Woodland, 
760-850 m., 2 Mar. 1910, Harris 10877 (GH ; K ; NY ; UCWI ; US). 

Forma 2. Leaf-blades varying from narrowly ovate to narrowly obovate, mostly 
more or less elliptic, 0-7-2-5 cm. long, 0-4-1 cm. broad. Corolla deep rose or red- 
purple (" deep lilac "). 

Jamaica : Trelawny : End of Crown Lands Road coming from Peckham, 300-600 m., 
Mar. 1954, Robertson 207g (UCWI). St. Elizabeth : Ipswich, 150-210 m., 7 Mar. 
1917, Harris 12308 (BM ; GH ; K ; NY ; UCWI ; US). Maggotty, 5 July 1928, 
Orcutt 2123 (BM). St. Ann : Moneague Hills, 1850, Alexander (K). Somerton 
district, 500 m., 28 Mar. 1956, Steam 608 (BM) ; Proctor iig2i (IJ). Manchester : 
Mandeville, Oct. 1886, Hart 1545 (K). Clarendon : Aenon Town to McKay, 700 m., 
Apr. 1962, Adams 11060 (UCWI), Adams 11062 (UCWI) ; Feb. 1961, Adams goo2 
(UCWI). Peckham Woods, 300-600 m., Mar. 1956, Robertson 2485 (UCWI). 

Forma 3. Leaf-blades mostly narrowly obovate, 1-2*5 *^^- ^^^%> O'4-i cm. broad. 
Corolla rose-pink, red-purple, bright rose or crimson. 

Jamaica : Trelawny : Island View Hill, Wilson Valley district, i| miles N. of 
Warsop, 600-670 m., 13 Oct. i960, Proctor 21407 (BM ; IJ) ; 30 Apr. 1964, Proctor 
24826 (BM ; IJ). Ramgoat Cave District, 450 m., 4 July 1955, Howard & Proctor 
14406 (BM ; IJ ; LS) ; 19 Jan. 1956, Howard, Proctor & Steam 14674 (A ; BM ; IJ). 
Norwood to Bailey, 770 m., Apr. 1963, Adams 12474 (UCWI), I247g (UCWI). Man- 
chester : Glasgow near Troy, 400 m., 18 Sept. 1906, Harris g478 (BM ; NY ; UCWI). 
Troy to Oxford, Sept. 1906, Britton 678 (NY). Clarendon : Peckham Woods, 760 m., 
26 May 1955, Proctor 10236 (BM ; I J) ; May 1955, Robertson ig85 (UCWI). St. 
Thomas : Between Whitehall and Big Hill, 300-900 m., 12 Feb. 1953, Proctor 7668 
(I J ; LS ; NY). Gorge of Plantain Garden River, 330 m., 15 Mar. 1956, Steam 
486 (A ; BM ; UCWI). 

3b. Oplonia armata var. pallidior Stearn, stat. & nom. nov. (Text-figs, i, 2 0, 
3 A, B; PI. 38). 

Anthacanthus acicularis sensu Lindau in Urban, Symb. Antill. 2 : 214 (1900) pro parte 

min., tantum quoad specim. Harris 6324 ; non (Swartz) Nees. 
Psilanthele minor Lindau in Urban, Symb. Antill. 7 : 383 (1912). 

Shrub or small tree to 6 m. high. Leaf-blades mostly ovate or elliptic, sometimes 
broadly elliptic or narrowly obovate, the base rounded to attenuate, 0-5-4 cm. long, 
0-3-2 cm. broad, the veining faint or not evident beneath. Corolla pale purple or 
lilac or almost white with purplish markings. 

Jamaica : Hanover : Green Island, Fish River, in coastal thicket, 15 Mar. 1908, 
Harris 10257 (BM ; F ; K ; NY ; US ; isotypes oi Psilanthele minor). Upper S. slope 



I 



OPLONIA AND PSILANTHELE (ACANTHACE AE) 303 



of Dolphin Head, 370-490 m., 11 Apr. 1955, Proctor 10018 (BM ; GH). Coastal 
thicket, Pedro Point, Mar. 1908, Britton 2184 (NY), Wooded limestone sea-cliffs 
I mile NW. of mouth of the Great River, 3-15 m., 15 Sept. 1957, Proctor 16655 (BM ; 
IJ). Lucea to Montego Bay, Mar. 1908, Britton 21^5 (NY). Westmoreland : Sea- 
coast, Negril, 10 Mar. 1908, Harris 10224 (NY ; UCWI ; US) ; Britton & Rollick 20^3 
(NY). Little Bay, 15 m., Dec. 1961, Adams ioi6g (UCWI). 5/. James : Glen Devon 
Road, Montego Bay, 1924, Norman 100 (BM). Hills overlooking Ironshore estate, 
60-240 m., 16 Feb. 1963, Proctor 23236 (IJ). 5/. Elizabeth : Coast about 2 miles E. 
of White House, 22 Feb. 1958, Yuncker 18012 (BM ; F ; S). Vicinity of Merriman's 
Point, in dry coastal woodland, 9 m., 10 May 1956, Steam 863 (A ; BM ; UCWI) ; 
Proctor 15337 (IJ). Near Fort Charles, 8 Jan. 1959, Robertson 53go (UCWI). Be- 
tween Southfield and Pedro Cross, 360 m., 9 May 1956, Steam 843 (BM). Lover's 
Leap, Santa Cruz Mountains, Sept. 1907, Britton 1151 (IJ). Yardley Chase, lime- 
stone cliff, 490 m., 4 Sept. 1907, Harris 9675 (NY ; UCWI ; US), i mile NW. of 
Alligator Pond, 14 Sept. 1954, Howard 6' Proctor 13807 (IJ). o-6 mile W. of Lititz, 
240 m., 21 Jan. 1956, Howard, Proctor & Steam 14740 (A ; BM ; IJ) ; Sept. 1907, 
Harris 9675 (NY ; US). Kaiser mine area S. of Gutters, 13 Sept. 1954, Howard & 
Proctor 13760 (I J ; NY). Manchester : Alligator Pond, i May 1896, Harris 6324 
(NY ; UCWI). Near Great Bay, 3 miles SE. of Alligator Pond, 15 m., 3 Apr. 1956, 
Steam 641 (BM). i mile ESE. of Gods Well, on N. side of Round Hill, 90 m., 3 Apr. 
1956, Proctor 11942 (BM ; IJ). St. Ann : Dry coastal thorn forest near the sea, 
Reynolds bucket line, 20-21 Dec. 1953, Howard & Proctor 13577 (U)- Fort Point, 
Dry Harbour (Discovery Bay), 3 m., 12 Apr. 1956, Steam 726 (BM). i mile W. of 
Discovery Bay, 30 m., Apr. i960, Adams 6855 (UCWI). Clarendon : E. end of Port- 
land Ridge, 15 m., 18 Mar. 1943, Lewis 104 (IJ). Round Hill near Milk Pen, 45 m.. 
May 1962, Adams 11143 (UCWI), 11147 (UCWI). Milk River, 3-6 m., 5 June 1965, 
Proctor 26894 (BM). 

0. armata as accepted here includes plants of such diversity in habit, size and shape 
of leaf, and colour of corolla that Lindau referred them to three species and two 
genera. The numerous specimens collected in Jamaica since Lindau's publications 
have increased rather than diminished the difficulty of classifying them. Swartz's 
original diagnosis of his Justicia armata reads : " I. fruticosa, aculeata, foliis oblongis 
emarginatis coriaceis nitidis. Jamaica ". The description in his Flora Indiae 
Occidentalis attributes to it " Caulis fruticosus, 3-4 pedalis . . . Folia breviter petio- 
lata cuneato-obovata, emarginata, nervosa . . . Aculei gemini . . . spinaeformes . . . 
Flores axillares, solitarii, pedunculati, coccinei " and gives its habitat as " in rupibus 
montium Jamaicae septentrionalis ". Red-flowered plants of this character occur 
in the interior of Jamaica, e.g. at White Rock Hill, which may possibly be the type- 
locality, as it stands by a route between Montego Bay and Savanna La Mar ; Swartz 
travelled across the island from the one to the other during his stay in Jamaica (cf . 
Steam, 1965). The number of flowers may vary up to four in a fascicle. Leaves of 
such red-flowered plants vary in size from 0-2 cm. to 2-5 cm. long, and display aston- 
ishing diversity in shape : three main types, described above as forms 1-3, may be 
distinguished under var. armata. Except in colour of corolla, it does not seem 
possible to distinguish from these a very variable assemblage of plants, here put 



304 A SURVEY OF THE TROPICAL GENERA 

under var, pallidior, occurring mostly at or near the coast and having pale purple, 
lavender-blue or almost white flowers ; one of these, in a spineless state, was described 
by Lindau in 1912 as Psilanthele minor, the specific epithet of which is certainly in- 
appropriate for a variant of 0. armata forming " a small tree up to 20 feet high in 
coastal thickets ", as noted by Harris for the type-collection {Harris 102 ^y) ; Harris 
recorded it as having " flowers pale blue ". This, however, is also evidently allied 
closely to plants inhabiting the interior of Jamaica and possessing larger leaves and 
more numerous flowers (4-15 in a fascicle) described by Lindau in 1908 as P. jamai- 
censis. The latter occurs in at least one locality, White Rock, with 0. armata var. 
armata f . i ; according to the limited material from this place, they here keep distinct, 
the difference in flower colour probably serving to prevent cross-pollination. 

Swartz's description of his Justicia armata as having sohtary red flowers excludes 
0. jamaicensis. A specimen presented by him to Sir Joseph Banks as his /. armata 
unfortunately lacks corollas. It has glabrous shoots, spines about 5 mm. long, 
solitary flowers with pedicels to i cm. long and calyx about 5 mm. long, but oblan- 
ceolate or narrowly obovate leaves, varying from 2-5 cm. long, i cm. broad to 5-5 cm. 
long, 2 cm. broad. No specimen exactly corresponding to this has been seen among 
the numerous collections now available from Jamaica. In its solitary flowers it 
agrees with 0. armata as accepted here on the basis of Swartz's description ; in its 
large leaves it comes closer, however, to 0. jamaicensis. It might possibly represent 
a hybrid between them. Its existence suggests that 0. jamaicensis could be united 
with 0. armata, but in view of the distinctness of all the other specimens they are 
here kept apart. In Swartz's herbarium at the Riksherbarium, Stockholm, there 
are four pieces labelled " armata ", evidently part of the same gathering, one piece 
with a spine to i cm. long, another with four pedicels arising from the same axil. This 
agrees with 0. jamaicensis. The descriptions in Swartz's Flora Indiae Occidentalis 
for the most part are evidently based on notes made directly from living plants in 
Jamaica and, as far as they go, are quite accurate. The explanation of the discrep- 
ancy between Swartz's description and his now available material may be that he 
described his armata from one living plant and kept dried specimens from another. 

4. Oplonia spinosa (Jacq.) Raf., Fl. Tellur. 4 : 65 (1838). (Text-figs. 2 A, B, C ; 
PI. 43a). 

Justicia spinosa Jacq., Enum. Syst. PI. Carib. : 11 (1760) ; Select. Stirp. Amer. : 2, t.2 fig. i 
(1763) ; Select. Stirp. Amer. Ic. Pict. : 7, t.2 (1780).— L., Sp. PL, ed. 2, 2 : 1663 (1763) 
excl. pi. Milleri. — Swartz, Obs. Bot. : 12 (1791). — ^Willd. in L., Sp. PL, ed. 4, i : 100 (1797)- 
— Vahl, Enum. PL i : 167 (1804). 

Eranthemum spinosum (Jacq.) R.Br, ex Roem. & Schult. in L., Syst. Veg., ed. nov. i : 176 
(1817). 

Eranthemum emarginatum Link in Jahrb. Gewachsk. i (3) : 47 (1820). — Schult., Mant. 
I : 154 (1822). 

Anthacanthus spinosus (Jacq.) Nees in DC, Prodr. 11 : 460 (1847). — Lindau in Urban, 
Symb. Antill. 2 : 216 (1900) excl. parte. — Britton & Millsp., Bahama Fl. : 403 (1920). — 
Urban, Symb. Antill. 8 : 655 (1921). — Britton & Wils. in Sci. Surv. Porto Rico & Virgin 
Is. 6 : 212 (1925). — Alain in Le6n & Alain, Fl. Cuba 4 : 500 (1957). — P. S. Green in Kew 
Bull. 23 : 275 (1969). 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 305 

Anthacanthus spinosus var. horridus Nees, loc. cit. (1847). 
Anthacanthus emarginatus (Link) Nees, torn. cit. : 461 (1847). 
Jasminum coeruleum Kuntze, Revis. Gen. PI. 2 : 410 (1891). 
Hesperanthemum spinosum (Jacq.) Kuntze, torn. cit. : 490 (1891). 

Shrub of variable habit, usually erect, sometimes dwarf but up to 3 m. (9 ft.) high : 
branches ascending, slender, glabrous on 2 opposite sides and pubescent on the 2 
intervening sides with short spreading hairs or densely and minutely pubescent all 
round, sometimes with the hairs on two sides slightly longer ; spines curved or 
directed downwards slightly or strongly, sometimes almost straight and horizontal, 
4-12 mm. long, at base c.o-5-o-7 mm. thick, glabrous or minutely pubescent. Leaves 
short-stalked, the blade very variable in shape and size, mostly narrowly elliptic or 
narrowly obovate, sometimes broadly ovate or elliptic, the apex obtuse or rounded 
and almost truncate, entire or emarginate, the base cuneate, c.8-20 mm. long 
(including the 1-1-5 mm. long petiole), 3-9 mm. broad, glabrous, the veining below 
faint or evident. Flowers solitary or 2-6 in sessile axillary fascicles ; pedicels 
3-9 mm. long. Calyx 3-5 mm. long, the segments narrowly triangular or subulate. 
Corolla bluish purple or lavender, glabrous or pubescent, the tube c. 6-9 mm. long, 
the segments 6-7 mm. long. Stamens included or exserted, the filaments 1-5-2 mm. 
or 6 mm. long ; anthers c.i-8 mm. long. Style 9-10 mm. or 6 mm. long. Capsule 
CI -2 cm. long. 

Bahamas, from Grand Bahama, New Providence and Andros, over eastern Cuba, 
Hispaniola and Puerto Rico, to the Virgin Islands, mostly on coastal limestone, but 
ascending to 900 m. 

Bahamas : Grand Bahama : Barnett's Point, 5-13 Feb. 1905, Britton & Millspaugh 
2711 (F). New Providence : 13 Mar. 1888, Eggers 4487 (BM : K). Near Nassau, 
24 Mar. 1903, Curtiss 133 (BM ; F ; GH ; K) ; 8 Mar. 1905, Wight 174 (GH ; K). 
Point Montague, 27 Aug. 1904, Britton & Brace 311 (F). Rifle Range, 3 Feb. 1905, 
E. G. Britton 32yg (F). Andros : Mangrove Cay, Aug.-Sept. 1906, Brace 4g26 (F) ; 
16-19 Jan. 1910, Small & Carter 84^1 (F ; K). Fresh Creek, 10 June 1890, Northrop 
& Northrop 3g6 (F ; GH) ; 12 Sept. 1937, Matthews 131 (K). Nicols Town, 8 Apr. 
1890, Northrop & Northrop 396 A (K). Eleuthera : fide Britton & Millspaugh, loc. cit. 
(1920). San Salvador {Watling Island) : Graham Harbour to Cockburn Town, 
15 Mar. 1907, Britton & Millspaugh 6ig2 (F). Cockburn Town and vicinity, 12-13 
Mar. 1907, Britton & Millspaugh 6ogg (F). Rum Cay : Sclater's Hill, 4 Dec. 1905, Brace 
3g4g (F). Port Nelson, 23 Mar. 1966, Gillis 62g8 (A). Exuma : Cay N. of 
Wide Opening, 18 Feb. 1905, Britton & Millspaugh 2^83 (F). Crooked Island : 
Landrail Point, Jan. 1906, Brace 4560 (F). Long Cay : Dec. 1905, Brace 4108 (F.). 
Acklins Island : Spring Point, Dec. 1905-Jan. 1906, Brace 4281 (F.) Caicos Islands : 
Bellemont, North Caicos, 2 Mar. 191 1, Millspaugh & Millspaugh gi88 (F). Jackson- 
ville, East Caicos, 26 Feb. 1911, Millspaugh & Millspaugh go83 (F). Vicinity of 
Cockburn Harbour, South Caicos, 17 June 1954, Proctor 8830 (IJ). Great Inagua 
Island : Between Matthew Town and Lighthouse, 14 Oct. 1904, Nash & Taylor 1053 
(F ; K). Salt Pond Hill, 12 Oct. 1904, Nash & Taylor gg3 (F ; K). 

Cuba : Oriente : Mayari, Charrascos cerca de la crestas del Cristal, 2-7 Apr. 1956, 
Alain, Acuna & Lopez Figueiras syoi (LS), 570J (LS), 5yo8 (LS), 5810 (LS). Cerro 



3o6 A SURVEY OF THE TROPICAL GENERA 

de Miraflores, Cananova, July 1942, Leon & Mentero 21152 (LS). Moa, lieux 
rocheux, 16-23 Apr. 1943, Marie-Victorin, Clemente & Alain 21485 (LS ; MX), 
2J77<§ (LS). Moa, Chemins de Cayo Chiquito, 18 May 1944, Clemente 3646 (LS ; MT). 
Moa, bois en arriere de la Mine Franklin, 20 May 1944, Clemente 3665 (LS ; LT). Moa, 
Playa la Vaca, Apr. 1943, Marie-Victorin, Clemente & Alain 21503 (LS). Rio 
Yagrumaje, 16-23 Apr. 1943, Marie-Victorin, Clemente & Alain 21621 (LS), 21625 
(MT). Moa, chemin de la mine Delta, a Test de la vallee du Rio Cayoguan, Marie- 
Victorin & Clemente 2iyy8 (LS ; MT). Moa, Orillas del Rio Cayoguan, 13 July 1949, 
Clemente & Alain 880 (LS). Moa, curso del Rio Cayoguan, 13 July 1949, Alain & 
Clemente 856 (LS). Sierra de Nipe, 24 July 1914, Ekman 2iyg (S). Sierra de Nipe, 
headwaters of Rio Piloto, 750 m., 15 May 1915, Ekman 5yoi (S) ; 375 m., 3 June 1924, 
Ekman giyo (S). Headwaters of Brazo Dolores, 925 m., 15 Oct. 19 14, Ekman 31 ly 
(S). Charrasco, Pefia Prieta, Magdalena, 30 July 1953, Alain 34gi (LS). 

Haiti : Without precise locality, Poiteau (BM ; P) ; Schomburgk loy (BM). 
Dipartement du Nord Quest : Vicinity of Jean Rabel, 1-13 Mar. 1929, Leonard & 
Leonard 13674 (A ; GH). Vicinity of Port a I'Ecu, Mar. 1929, Leonard & Leonard 
13876 (K). Dipartement de I'Artibonite : Plaisance, Gonaives, 400 m., May 1900, 
Buck 375 (IJ). Departementde I'Ouest : Duchene, Port au Prince, 800 m.. Mar. 1920, 
Buck 1851 (IJ). Massif de la Selle, Port au Prince, 300 m., 21 Apr. 1926, Ekman 
5g23 (K). Massif de la Hotte, gr. Morne Rochelois, Miragoane, 25 Mar. 1927, 
Ekman y8g6 (S). 

Dominican republic : Without precise locality, 1852, Schomburgk (BM). Puerto 
Playa, sea-level, i May 1959, Jimenez 3g3i (BM), Opposite San Gabriel, S. side of 
Samana Bay, sea-level, 10 April 1921, Abbott I2g3 (GH). Barahona, Enriquillo 
(Petit Trou), sea-level to 100 m., 15 Feb. 1922, Abbott 1756 (GH). 

Puerto Rico : Aquadilla: Cape Borinquen, 2 Mar. 1929, Britton & Britton g283 
(F ; GH). Quebradillas, 2 Mar. 1914, Britton, Britton & Cowell ig58 (F). Bayamon : 
Hills near Bayamon, 20 m., 23 Feb. 1900, Heller 4688 (A ; F ; GH). Humacao : Punto 
Cabalero, 29 Sept, 1886, Sintenis 5202 (BM ; GH). Arecibo : Prope Utuado ad Los 
Angelos, 17 Jan. 1887, Sintenis 5gi3 (K). Vieques Island : Cayo Puerto Real, 
7, Feb. 1914, Shafer 2753 (GH). 

St. Thomas : Without precise locality. Herb. Willdenow 2y6 (B, type of Eranthe- 
mum emarginatum ; BM, photo) ; Feb. 1874, Kuntze (K, isotype of Jasminum 
coeruleum) ; Feb. 1887, Eggers 630 (A ; P). Cowells Hill, Jan. 1881, Eggers 203 (K). 

St, John : Without precise locality, i March 1957, Robertson 26 (A). 

Virgin Gorda : North Sound, 29 Oct. 1918, Fishlock 6 (A ; GH). Eastern slope 
of Virgin Peak, 150-420 m., 15 Apr. 1956, A. C. Smith 10562 (A ; K ; IJ). 

St. Croix : Without precise locality, L. C. Richard (P) ; May 1896, Ricksecker 335 
(F ; GH). E. slope of Mount Eagle, 230 m., 21 Dec. 1959, Proctor 20448 (A ; BM ; 
IJ). Near Kingshill, 20 Jan. 1961, Howard 15658 (A). 

Jacquin's epithet spinosa is the earliest one given to a member of the group ; his 
original diagnosis under Justicia spinosa being but three words long, " Justicia 
monanthera, spinosa ", this epithet has excusably been applied to a diversity of 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 307 

plants. The term " monanthera " as used here means that the two loculi (thecae) 
of the anther are at the same level and opposite one another, thus obviously being 
part of one anther, in contrast to the term " dianthera " referring to stamens in which 
the two loculi are at different heights, the filament thus appearing to possess two 
anthers. Jacquin's diagnosis served to distinguish his species from all the species 
until then referred to Justicia ; it fits, however, almost all the members of the genus 
Oplonia and its appUcation would be obscure indeed but for the detailed description 
and the illustration pubHshed three years later in his Selectarum Stirpium American- 
arum Historia. This estabUshes the name as based on a shrub up to 1-5 m. high with 
more or less elliptic or narrowly ovate leaves about 1-5 cm, long and cuneate at base, 
with spreading spines shorter than the leaves, one to four flowers at an axil, a short 
calyx only half the length of the corolla-tube, and a purple corolla, its habitat being 
" in fruticetis & arbustis maritimis circa Portum Principis in Domingo ". There 
can be no doubt that a plant collected at Duchene, Port au Prince, Haiti, by W. Buch 
[Buck 1851 in I J) represents Jacquin's Justicia spinosa. While some spines on this 
specimen spread horizontally, most of them curve markedly downwards and they 
vary from 8 to 15 mm. in length. The leaves range from i to 2 cm. in length, 2-5 to 
7 mm. in width. There does not seem any justification for separating specifically 
from this gathering other Haitian specimens with downwards curved spines, e.g. 
Buch sy^ from Gonaives, which has, however, small ovate leaves rounded at base, and 
Leonard & Leonard 13624 from Jean Rabel, which has lanceolate, narrowly elliptic 
and narrowly oblanceolate leaves. Buch's Duchene material links up, however, with 
comparatively large-leaved specimens from the Moa district of Cuba (e.g. Marie- 
Victorin, Clemente & Alain 2iyy8) on which the leaves vary from elliptic to ovate and 
narrowly obovate and from 7 mm. to 35 mm. long. The range of variation within a 
a series of specimens collected in the Moa district does not make it feasible to separate 
specifically the plant from St. Thomas on which the names Eranthemum emarginatum 
Link and Jasminum coeruleum Kuntze are based. Leonard and Leonard's small- 
leaved material from Jean Rabel in turn links up with the small-leaved plants 
characteristic of the Bahamas. There is indeed surprisingly little difference between 
some specimens from New Providence, Bahamas (e.g. Curtiss 133) and some from 
St. John, Virgin Islands (e.g. Robertson 26). 

5. Oplonia microphylla (Lam.) Steam, comb. nov. (Text-figs. 2 J-M, 4 ; PI. 44). 

Justicia microphylla Lam. in Tab. Encycl. M6th., Bot. i : 37 (1791). — Vahl, Eclog. Amer. 

2 : I (1798) ; Enum. PL i : 168 (1804). 
Eranthemum microphyllum (Lam.) R.Br, ex Roem. & Schult. in L., Syst. Veg., ed. nov. i : 

176 (1817). 
Eranthemum lycioides Link in Jahrb. Gewachsk. i (3) : 47 (1820). 
Anthacanthus microphyllus (Lam.) Nees in DC, Prodr. 11 : 461 (1847). — Griseb., Fl. Brit. W. 

Ind. Is. : 457 (1862). — Duss in Ann. Inst. Colon. Marseille 3 : 427 (1897). 
Anthacanthus lycioides (Link) Nees, loc. cit. (1847). 
Anthacanthus cuneatus Nees, loc. cit. (1847). 

Hesperanthemum microphyllum (Lam.) Kuntze, Revis. Gen. PI. 2 : 490 (1891). 
Anthacanthus acicularis sensu Lindau in Urban, Symb. Antill. 2 : 214 (1900) pro parte ; 

non (Swartz) Nees. 

D 



3o8 A SURVEY OF THE TROPICAL GENERA 



I 



Erect much-branched spiny shrub to 1-5 m. high ; branches spreading almost 
horizontally or ascending, slender, brownish or grey when young, glabrous on 

2 opposite sides and pubescent on the 2 alternating sides, with short or moderately 
long spreading or ascending whitish hairs o- 1-0-3 rn^i- long, or pubescent all round, 
the internodes 3-15 mm. long ; spines ascending or almost horizontally spreading, 
straight 3-12 mm. long, glabrous or pubescent. Leaves short-stalked, the blade 
elliptic or narrowly elliptic or narrowly ovate to oblanceolate or narrowly obovate, 
the apex rounded or almost truncate, slightly emarginate and mucronate, the base 
cuneate, c.3-12 mm. long (including the 0-5-1 mm. long petiole), 1-5-6 mm. broad, 
glabrous above, glabrous or occasionally pubescent below along the midrib, the 
veining faint not evident. Flowers soUtary or paired ; pedicels 2-7 mm. long, 
glabrous. Calyx 2-3-5 ^^- lorig. the segments narrowly lanceolate. Corolla pale 
purple, lilac or almost white with purple dots on the lower lip, minutely pubescent, 
the tube c.6-7 mm. long, the segments 5-6 mm. long. Stamens almost included or 
long-exserted, the filaments c.i-5 mm. or 4-5 mm. long, the anthers 1-5 mm. long. 
Capsule c.i-2 cm, long ; seeds c.2 mm. long. 

South-eastern Jamaica, Hispaniola and Puerto Rico, Lesser Antilles south to the 
Grenadines, at sea-level to about 600 m., mostly under 200 m. 

Jamaica : Without precise locality, 1730, Houstoun (BM) ; W. Wright (BM) ; 
Macfadyen (K, t5^e of Anthacanthus cuneatus). St. Catherine : Port Henderson 
Hills, Nov. 1954, Asprey 20^g (UCWI). Lazaretto, above Port Henderson, 30-75 m., 

27 Mar. 1955, Proctor ggyi (BM ; IJ). Kingston & St. Andrew : E. of Militia 
Rifle Range, 45 m., 8 Jan. 1904, Harris 86og (NY). Long Mountain, S. side, 90 m., 

3 July 1907, Harris g6o6 (BM ; K ; NY ; UCWI) ; 90-220 m., 19 June 1959, Webster, 
Ellis & Miller 8364 (BM). W. slope of Long Mountain, along road to Wareika, 
75-370 m., 7 May 1954, Farr 5133 (IJ) ; 140 m., 3 Jan. 1956, Steam 66 (A ; BM ; 
UCWI) ; Proctor 11436 (I J) ; 150 m., 7 May 1956, Steam 830 (A ; BM). Gorge of the 
Cane River, 75-120 m., 22 May 1955, Proctor loigs (I J) ; 29 Oct. 1957, Yuncker 17215 
(F). St. Thomas : Four Mile Wood, between Bull Bay and Grant's Pen, 150 m., 

28 Nov. ig^8, Proctor 18268 {A ; BM ; IJ). Grant's Pen, 6 m., Dec. 1959, Adams 
5665 (UCWI). Yallahs, 30 m., 16 June 1952, Araque-Molina & Barkley 22J448 (BM ; 
IJ). Vicinity of Gypsum mines, i mile N. of Eleven Mile, 150 m., 8 Nov. 1963, 
Proctor 24186 {BW). 

Haiti : Departement de I'Artibonite : Andr^, Gonaives, 200 m., Apr. 1899, Buch 144 
(IJ). Departement de l' Quest : Plaine Cul-de-Sac, near Gauthier, 24 Feb. 1925, 
Ekman 3321 (GH ; F ; K ; S). Departement du Sud : Prope Port k Piment, Aug. 
1917, Ekman 6g8 (A ; S). Inter Cahauane et Aux Anglais, July 1917, Ekman 3y8 (S). 

Dominican republic : Without precise locality, 1852, Schomburgk 57 (BM). 
Barahona : Barahona, Las Salinas, 40 m.. May 191 1, Fuertes 1105 (A ; P). Santiago : 
distr. of San Jose de las Matas, 600 m., 12 Oct. 1931, Valeur ^24 (A ; BM ; F ; K ; S). 
Pico Diego de Ocampo, 1220 m., 3 Mar. 1957, Jimenez 3455 (BM). 

Puerto Rico : Ponce : Ponce to Playa Las Cucharas, 13 Mar. 1913, Britton, Britton 
& Marble ig48 (A ; F). Ponce to Penuelas, 22 Mar. 1906, Britton & Cornell I2yi (F). 
Culebra : Mar. 1906, Britton & Wheeler 23 (F). Vieques : Jan. 1876, Eggers (GH). 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 309 

St. Croix : Without precise locality, L. C. Richard (P, isotype of Justicia micro- 
^hylla). 

ToRTOLA : Coastal hill, 1918, Fishlock 10$ (K). 

St. Barthelemy : Morne de Grand Fond, 250 m., 11 Oct. 1951, Le Gallo 311 (A ; 
.S ; NY). Grand Platon, 150-200 m., 30 Oct. 1952, Le Gallo lo^y (LS ; NY) ; 
E2 Nov. 1953, Le Gallo 2180 (A). 

Antigua : Without precise locaUty, Wullschlagel 452 (GOET) ; Nicholson (K). 
Santa Maria Hill, 25 Apr. 1937, Box y^y (A ; BM). Little Cove near Half Moon Bay, 
July i960, D. R. Harris 34 (BM). 

Guadeloupe : La Desirade : Duchaissang 461 (GOET) ; 10-150 m., Nov. 1935, 
Stehle424 (I J) ; 100 m., Stehle & Stehle yogi (A) ; 8 June i960. Proctor 21212 (A). 

Grenadines : Petit St. Vincent : East end, Mar.-Apr. 1950, Howard 10916 (A ; 
BM). Little Martinique : Mar. 1950, Howard iog33 (A ; GH). 

0. microphylla was published by Lamarck in 1791 under the name Justicia micro- 
phylla with an extremely brief diagnosis and a vague indication of its provenance as 
follows : "J. fruticosa, foliis obovatis coriaceis, spinis axillaribus brevissimis, 
pedunculis unifloris lateralibus. Ex insulis Caribaeis. Richard, an justicia armata 
Swartz. " The type-specimen in Lamarck's herbarium, now at the Museum National 
d'Histoire Naturelle, Paris, gives no more precise information about its place of 
collection but fortunately it matches so exactly a localized specimen, under another 
name, in L. C. Richard's own herbarium, likewise at Paris, that they presumably 
formed part of the same gathering made " in fruticetis sepibus Stae Crucis ", i.e. on 
the island of St. Croix, Virgin Islands, which Richard visited between February 1786 
and November 1787 (cf. Urban, Symb. Antill. 3 : iii (1902) ). Richard's material 
has light brown branches which are almost glabrous on two opposite sides but 
pubescent on the other two sides with moderately long whitish hairs to 0-3 mm. long ; 
the short stout slightly ascending spines are up to 3 mm. long ; its numerous, narrowly 
obovate leaves are 4-10 mm. long, 3-6 mm. broad, and sometimes have a few hairs 
along the midrib below. The small leaves and open habit, with branches spreading 
laterally from the main stem at an angle of about 6o°-8o°, give it a distinctive 
appearance which it shares with other specimens from the Lesser Antilles ; these 
diverge from it principally in having somewhat longer (5-12 mm. long) spines. On 
the dry coastal limestone of south-eastern Jamaica grow similar small-leaved plants 
with, however, more erect branches which ascend from the main stem at an angle of 
about 50° or less. There is much variation in degree of hairiness within this Jamaican 
population. Anthacanthus cuneatus was founded on a gathering by James Macfadyen 
with markedly hairy shoots, spines and leaves ; among modern Jamaican collections 
this is represented by Araque-Molina & Barkley 22J448 from Yallahs, St. Thomas. 
It does not seem possible, however, to make any satisfactory distinction between this 
and other less hairy Jamaican specimens of the same area : distinctions between 
specimens from Jamaica and those from Hispaniola and Puerto Rico are likewise not 
clear-cut. Hence the epithet microphylla is here used to cover them aU. 



3IO 



A SURVEY OF THE TROPICAL GENERA 



6. Oplonia tetrasticha (C. Wright ex Griseb.) Stearn, comb. nov. 

Anthacanthus tetrastichus C. Wright ex Griseb., Cat. PL Cub. : 198 (1866). — Alain in Le6n 

& Alain, Fl. Cuba 4 : 500, fig. 219 (1957). 
Hesperanthemum tetrastichum (C. Wright ex Griseb.) Kuntze, Revis. Gen. PL 2 : 490 (1891). 
Anthacanthus acicularis sensu Lindau in Urban, Symb. Antill. 2 : 214 (1900) pro parte min. ; 

non (Swartz) Nees. 

Much branched spiny shrub, up to about 2 m. high ; branches stout, yellowish when] 
young and pubescent all round or only on two opposite sides with short spreading! 
hairs, the internodes 5-15 mm. long ; spines horizontally spreading or very slightlyj 
ascending straight, rigid, 1-3-5 cm. long, to i mm. thick, simple (var. tetrasticha) 01 
branched (var. polyece), glabrous or minutely pubescent. Leaves short-stalked, the! 
blade narrowly obovate to narrowly elliptic, the apex rounded, sHghtly mucronate,] 
the base cuneate, 6-10 mm. long (including the 1-3 mm. long petiole), 2-3 mm. broac 
much shorter than the spines, glabrous, the veining not evident. Flowers solitarj 
in the axils ; pedicels 4-6 mm, long, glabrous. Calyx c.3-5 mm. long, the segments 
subulate. Corolla lilac or "bright violet with white purple-dotted patch in throat 
(fide Proctor), glabrous or pubescent, the tube 4-5-10 mm. long, the segments 
c.3-5 i^i^- loiig- Stamens included or exserted, the filaments c.i or 4 mm. long, the 
anthers 0.1-5 mm. long. 

Endemic to Cuba. 



6a. Oplonia tetrasticha var. tetrasticha (Text-figs. 2 D, E). 
Spines all simple, not branched. 

Cuba : Without precise locality : 1860-64, C. Wright 3067 (BM ; K ; S, isotypes). 
Pinar del Rio : Cuabales al Este de la Loma de Cajalbana, La Palma, 10 June I950,j 
Alain & Acuiia ijg^ (LS) ; 3 Apr. 1954, Alain jgog (LS). Norte del Pan de Guajaib6n^ 
5 Apr. 1952, Alain 242 j (LS). Habana : Guanabacoa, Lomas de las Jatas, 3 June 
1923, Ekman 16530 (S). Cojimar, 24 May 1914, Ekman iioi (S). ColumbiaJ 
Habana, 22 June 1912, Leon 3326 (LS ; P). Valle del Almendares, Dec. 1908, Leof 
465 (LS). Loma S. de la Jata, Guanabacoa, 4 Jan. 1912, Leon 2865 (LS). Loma de 
Machado, Jibacoa, 2 Jan. 1929, Leon & Roig 13756 (LS). Matanzas : CamariocaJ 
1849, R'^g^^ 24g (K). Cerca de la Boca de Canasi, 17 Oct. 1928, Leon I36g6 (LS)j 
Las Villas {Santa Clara) : Sabana de Motembo, 2 Jan. 1919, Leon &Fortem S^yg (LS)| 
Loma de Belen, cerca de Santa Clara, 28 June 1931, Leon I4g47 (LS), Cienfuego^ 
Bay, dry east-facing slopes of Loma San Juan, 11 July 1947, Wood & Atchison y4i( 
(IJ). Gavilan, Cienfuegos, 13 Mar. 1928, Jack 5815 (A ; LS). Alto de Farallon," 
Ensenada de Guajimico, S. de Cienfuegos, 21 Mar. 1953, Alain, Clemente & Wal- 
singham 283g (LS). Guajimico Bay, on coast S.E. of Cienfuegos, 3-15 m., 23 Mar. 
1957, Proctor 16430 (IJ). Oriente : El Cobre,60ct. 1916, Ekman 7832 (S). Santiago, 
in limestone hills W. of Playa Siboney, 8 Nov. 1917, Ekman 8723 (S). Santiago, pr. 
Aquadores, 1^22, Ekman 15631 (S). Sierra de Nipe, between Rio Bayate and Rio 
Piedra, 8 Sept, 1922, Ekman 15042 (BM ; S), 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 311 

6b. Oplonia tetrasticha var. polyece Stearn, var. nov. (Text-figs, 2 F, G.) 

Spinae primariae rectae lignosae 1-3-5 cm. longae plerumque ramosae spinas 
secundarias plerumque 2 interdum 4 geminatas 1-12 mm. longas vel tantum gemmas 
minutas geminatas ferentes. 

Primary spines straight, 1-3-5 cm, long, usually branched and bearing 2-4 paired 
spines 1-12 mm. long or in their place minute paired buds. 

Cuba : Oriente : Sierra Maestra, on the edge of Arroyo Corojo, a tributary of Yara, 
near Nagua, c.150 m., 5 Aug. 1922, Ekman 14735 (BM). Cuabales de serpentina, 
Holguin, May 1939, Leon & Carahia i8g8i (LS) . Pinal Mayari, Sierra de Nipe, 7 Apr. 
1941, Leon, Marie-Victorin, Clemente & Alain igg20 (LS). Rente, Bahia de San- 
tiago, I May 1943, Clemente 2680 (LS). Club de Punta Gorda, Bahia de Santiago de 
Cuba, I Mar. 1953, Lopez Figueiras gii (LS). Manigua costera de Santiago de Cuba, 
Mar. 1954, Lopez Figueiras 1131 (LS). Antigua linea de Firmeza, Santiago de Cuba, 
May 1946, Clemente 5014 (LS). Manigua de Ciudamar, June 1946, Clemente sojg 
(BM ; LS). Camino viejo del Morro, Oct. 1947, Clemente 5657 (BM, holotypus ; LS) ; 
July 1943, Clemente 2go6 (LS). Manigua de Aguadores, Santiago de Cuba, May 1946, 
Clemente 5021 (LS). Punta de la Gaviota junto a la Playa de Siboney, Santiago de 
Cuba, 24 Feb. 1952, Lopez Figueiras j^g (LS). Farallon de la Perla, 14 Feb, 1911, 
Shafer 8yy2 (GH ; K). La Perla, prope villam Monte Verde, 1859, C. Wright jOg 
(GH ; K). Las Ninfas, Guantanamo, Feb. 1919, Hioram 2354 (LS). Macambo, Via 
Azul, 28 Dec. 1954, Alain & Lopez Figueiras 4210 (LS). Charrascos de Jauco, July 
1924, Leon iigi (LS). Region de Baracoa, entre Jauco y Cajobabo, July-Aug. 1924, 
Leon iig8y (LS), i2oyg (LS), 12031 (LS). 

The long and rigid sharply pointed spines projecting from the branches in four 
ranks (whence the epithet tetrasticha) have earned this species the Cuban vernacular 
name " no-me-toques ", They are longer than in any other species of Oplonia. The 
type collected by Charles Wright somewhere in western Cuba has simple stiletto-like 
spines, as have other specimens collected outside of the Oriente province. Within 
that province, along the south coast, there is a marked tendency for these major 
spines to bear a pair or two pairs of subsidiary spines which are usually short but 
may be up to 12 mm, long. Numerous specimens with such branched spines have 
been collected around Santiago de Cuba, Many of them have the shoots hairy on 
two opposite sides only, the two sides alternating with these being glabrous, but some 
(e.g. Lopez Figueiras 35g from Siboney, near Santiago de Cuba) are hairy all round. 
The leaves of main shoots vary from 5 to 12 mm, in length ; those on the spines are 
much smaller. The plant would appear to initiate a lateral branch, then halt its 
development after producing a pair or less often two pairs of short spines, then stop 
its development entirely by making it terminate as a sharp spine. The few corollas 
examined of this variant have been glabrous, whereas the few corollas examined of 
typical 0. tetrasticha have been minutely pubescent. A specimen of the Baracoa 
region, Oriente, collected by Leon {Leon 12031), appears at first to have long simple 
spines as in typical 0. tetrasticha, but these bear nevertheless a pair of minute buds 
corresponding to the secondary spines of other specimens collected in this area by 
Leon, The Oriente population with branched spines is accordingly here distin- 



312 A SURVEY OF THE TROPICAL GENERA 

guished as var. polyece, the epithet from tt-oAu- (" many ") and -i^/cry (" pointed ") 
being formed by analogy with ei5i7/c7^ '("well-pointed") dJidX^TTT-qKi) ("fine-pointed"). 
Usually the spines are more than 1-5 cm. long. A specimen from Habana province, 
collected by Ekman {Ekman iioi) at Cojmar, which it does not seem possible to refer 
to any species but 0. tetrasticha, nevertheless has straight slightly ascending spines 
with the smallest down to 3 mm., the longest to 2 cm. long. Another Ekman 
gathering [Ekman 15042) from Oriente province has more slender spines than is 
typical and approaches 0. microphylla. 

7. Oplonia acicularis (Swartz) Stearn, comb. nov. (Text-fig. 2 H, I). 

Justicia acicularis Swartz, Nov. Gen. & Sp. PI. : 13 (1788) ; Fl. Ind. Occ. i : 29 (1797). — 

Vahl, Enum. PI. i : 169 (1804). 
Eranthemum aciculare (Swartz) R. Br. ex Roem. & Schult. in L., Syst. Veg., ed. nov. i : 177 

(1817). 
Anthacanthus acicularis (Swartz) Nees in DC, Prodr. 11 : 460 (1847) excl. parte. 
Anthacanthus jamaicensis Griseb., Fl. Brit. W. Ind. Is. : 457 (1862), nom. superfl. 
Hesperanthemum aciculare (Swartz) Kuntze, Revis. Gen. PI. 2 : 490 (1891). 
Anthacanthus spinosus sensu Lindau in Urban, Symb. Antill. 2 : 216 (1900) pro parte min. ; 

non (Jacq.) Nees. 

Low spreading much branched spiny shrub, mostly under 30 cm. high ; branches 
slender, brownish when young and pubescent with minute upward-pointing hairs on 

2 or all sides ; spines very slender, 8-14 mm. long, c. 0-3 mm. thick. Leaves short- 
stalked, the blade ovate to narrowly lanceolate, the apex obtuse, mucronate, the base 
cuneate or rounded, 7-25 mm. long (including the c.i mm. long petiole), 4-10 mm. 
broad, glabrous, the veining not evident. Flowers solitary in the axils ; pedicels 
c.3-4 mm. long, minutely pubescent. Calyx 2-2-4 ^^- long, the segments very 
narrowly lanceolate or subulate, ciliate. Corolla white, glabrous, the tube cyUndrical 
0.4-6-5 mm. long, the limb spreading with segments oblong, c.4-6 mm. long, 2-5- 

3 mm. broad, obtuse. Stamens included, inserted on the upper third of the corolla- 
tube, the filaments c.o-8 mm. long, the anthers 1-4 mm. long. Capsule c.1-5 cm. 
long ; seeds c.2-5 mm. long, rugose. 

Endemic to Jamaica. 

Jamaica : Without precise locality : 1862, March (K) ; Swartz (S, type collection) ;&. 
Wilson 453 (GOET ; K). Westmoreland : Between Grange Hill and Glasgow, Mar.-m 
Apr. 1908, E. G. Britton 28g4 (NY). Hills near Bulstrode, 12 Mar. 1908, E. G. Britton 
28yy (K ; NY). Portland : Gorge of the Swift River at Eden, | mile N. of Paradise, 
30 m., 19 Mar. 1956, Stearn 550 (BM), Proctor 11866 (IJ). Navy Island, June 1897, 
Fredholm joy 3 (NY ; US). Friendship Valley district near Ecclesdown, 330-360 m., 
23 May 1967, Proctor 2yg20 (IJ). Priestman's River, Nov. 1886, Hart 1545 (K) ; 
3 m., 13 Aug. 1957, Proctor 16574 (I J) > 18 Mar. 1958, Yuncker 18421 (IJ). Law- 
rence Bottom, 370 m.. Mar. 1961, Adams 9258 (UCWI). St. Thomas : Big Level 
district, S.E. end of John Crow Mountains, 450-600 m., 16 Mar. 1956, Stearn 498 
(BM.), Proctor 11818 {!]). 

Although 0. acicularis has been recorded (as Anthacanthus acicularis) from many 
West Indian islands, through confusion with other species, it is, in fact, a Jamaican 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 313 

indemic characterized by low growth, slender needle-like spines, mostly ovate leaves 
id small white flowers which make it easily distinguishable in a living state from 
)ther Jamaican species. It often grows in somewhat shaded places. Swartz's 
)riginal diagnosis of his Justicia acicularis reads : "I. fruticosa, diffusa, spinosa, 
spinis setaceis, floribus pedunculatis axillaribus solitariis. Jamaica ". Later he 
supplemented this in his Flora Indiae Occidentalis with a detailed description, giving 
particulars such as " Fruticulus pedalis . . . Folia petiolata, opposita ovata minora, 
apice acumine brevissimo , . . Spinae oppositae, setaceae . . . Flores . . . parvi, 
albidi " etc., which together apply only to the Jamaican species described above. 
It has a remarkably disjunct distribution within the island, most specimens having 
been collected on the extreme east and a few in the west but none in the middle. 

8. Oplonia purpurascens (Griseb.) Stearn, comb. nov. (Text-fig. 2 P, Q.) 

Anthacanthus purpurascens Griseb., Cat. PI. Cub. : 197 (1866). — Alain in Le6n & Alain, Fl. 

Cuba 4 : 500 (1957)- 
Hesperanthemum purpurascens (Griseb.) Kuntze, Revis. Gen. PI. 2 : 490 (1891). 
Anthacanthus acicularis sensu Lindau in Urban, Symb. Antill. 2 : 214 (1900) pro parte min. ; 

non (Swartz) Nees. 

Shrub ; branches ascending, slender, purplish, glabrous on 2 opposite sides, 
minutely pubescent on the 2 other sides when young ; spines ascending at an angle of 
c.48-58°, slender, straight, 4-8 mm. long, at base 0-5 mm. thick. Leaves short- 
stalked, the blade obovate or oblanceolate, the apex obtuse, entire or slightly 
emarginate, the base narrowly cuneate, 6-15 mm. long, 1-5-6 mm. broad, glabrous, 
the veining scarcely evident ; petiole 1-3 mm. long. Flowers solitary or paired in the 
axils ; pedicels c. 1-1-5 i^^i- loi^g- Calyx c.3-4 mm. long, the segments very narrowly 
lanceolate or subulate. Corolla " pale purple ", minutely pubescent, the tube 
c.8-9 mm. long, the segments c.6 mm. long. Stamens included. 

Endemic to Cuba. 

Cuba : Without precise locality, 1860-64, C. Wright 3066 (BM ; K ; S, isotypes). 
Pinar del Rio : Sierra de Vinales, 8 June 1923, Ekman 16382 (BM). Mogote de la 
Bandera, Vifiales, 9 Mar. 1924, Ekman 18658 (LS). Ensenada de los Muertos, 
Mogote de la Bandera, Viilales, 29 Mar. 1953, Alain 2860 (LS). Ensenada de la 
Bandera, Mogote de la Bandera, Vifiales, Mar. 1953, Alain 28g2 (LS). 

0. purpurascens is a species characteristic of the mogotes (" haystack hills " or 
" Kegel-karst ") of north-western Cuba. The very short pedicels distinguish it from 
allied species. Possibly here belongs a specimen collected in the district of Cienfuegos, 
Santa Clara by R. Combs {Combs 56^ ; K). 

9. Oplonia nannophylla (Urban) Stearn, comb. nov. (Text-fig. 2 N, PI. 45). 

Anthacanthus nannophyllus Urban in Fedde, Repert. Sp. Nov. 20 : 350 (1924). 

Low spreading shrublet branched from the base, 4-8 cm. high ; branches spreading 
or arching, slender, minutely pubescent when young, the internodes 1-3 mm. long ; 
spines almost horizontally spreading or ascending, straight or very slightly curved, 



314 



A SURVEY OF THE TROPICAL GENERA 



slender, 2-5-11 mm. long, at base to 0-5 mm. thick. Leaves short-stalked, almost 
sessile, much shorter than the spines, the blade obovate or narrowly elliptic, the 
apex obtuse, the base broadly cuneate, 2-5 mm. long, 1-3 mm. broad, glabrous, the 
cystoliths numerous, the veining not evident, rigid ; petiole c.o-5 mm. long. Flowers 
unknown. 

Endemic to Cuba. 

Cuba : Pinar del Rio : Cuabales, Oeste de la Loma de Cajalbana, La Palma, 3 Apr. 
1954, Alain sgoi (LS) ; Feb. 1953, Alain 2j^y (LS). Hahana : prope Guanacaboa in 
Lomas de la Yatas, 3 June 1923, Ekman 16534 (S, type). Las Villas {Santa Clara) : 
Manigua, Sabana de Motembe, 28 Aug. 1922, Leon 1138$ (GH ; LS). 

0. nannophylla is an inconspicuous plant of extremely reduced habit and probably 
slow growth, with minute leaves and short internodes, which has not yet been 
collected in flower. It is apparently restricted to north Cuba between 80° and 84°W. 




Fig. 17. Leaves of Oplonia from Madagasgar. A, B, O. vincoides ; C, O. acuminata 

D. O. minor var. meridionalis. 



10. Oplonia acuminata Stearn, sp. nov. (Text-fig. 17 C) 

Frutex 1-3 m. altus, inermis ; ramuli hornotini in sicco atri glabri [vel pubescentes], 
annotini graciles pallidi, internodiis plerumque elongatis ad 9 cm. longis. Folia 
manifeste petiolata ; lamina ovata, apice acuminata, basi rotundata in petiolum 
ad I cm. longum abrupte angustata, c. 2-5 cm. longa, acumine 5-9 mm. longo inclusa, 
1-4 cm. lata, glabra [vel supra sparse subter magis praecipue secus venas puberula], 
cystolithis tantum sub lente visibilibus sed numerosis, papyracea, venis arcuatis 
conspicuis. Flores axillares 1-3 ; pedicelli 7-9 mm. longi, glabri. Calyx c.5 mm. 
longus, sparse puberulus [vel glaber], segmentis subulatis. Corolla alba, bilabiata ; 
tubus 0.4-5 iiini- longus, sparse puberulus ; labium superum ascendens c. 8 mm. 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 315 

longum, segmentis duobus supra medium coalitis quamobrem bilobatum lobis liberis 
0.3-5 ^ni- longis 2 mm. latis ; labium inferum descendens trifida, segmentis anguste 
ovatis acutis c. 8-9 mm. longis, 4 mm. latis ; segmenta extra intraque praecipue 
apicem versus puberula. Stamina in fauce affixa, exserta ; filamenta laevia, in typo 
5-5 mm. longa ; antherae purpureae, c.2 mm. longae ; staminodia minuta. Ovarium 
glabrum ; stylus puberulus, c.12 mm. longus. [Capsula 0.1*7-2 -5 cm. longa]. 

Shrub 1-3 m. high ; branches slender, glabrous or pubescent when young ; spines 
absent. Leaves short-stalked, the blade ovate, the apex acuminate, the base rounded 
and abruptly narrowed into the petiole, 2-5 cm. long (including the 5-9 mm. long 
elongated tip), 1-4 cm. broad, almost glabrous or else pubescent particularly 
below along the veins ; petiole 5-10 mm. long. Flowers 1-3 in the axils ; pedicels 
7-9 mm. long, glabrous. Calyx c.5 mm. long, sparsely puberulous or glabrous, the 
segments subulate. Corolla white, puberulous, the tube c.4 -5 mm. long, the segments 
8-9 mm. long, 4 mm. broad, the upper two united to above the middle. Stamens ex- 
serted, the filaments (on the type) 0.5-5 ^n^^- long, smooth, the anthers c.2 mm. long. 
Style puberulous. Capsule 1-7-2-5 cm. long. 

Endemic to western Madagascar. 

Madagascar : Environs de Majunga ; arbuste 1-3 m., fleur blanche a 2 levres, 
la superieure entiere, Finferieure trilobee ; Nov. 1914, Perrier de la Bathie i2go8 
(P, holotype). Vallee du Fiherenana, a 15-25 kil. en amont de Tulear ; arbuste 2 m., 
coroUe a levre inferieure blanche, levre superieure blanc-verdatre, antheres pour- 
pres ; vers 200 m.. Mar, 1934, Humbert 14357 (K). 

This is separated from 0. vincoides by its somewhat broader, ovate, distinctly 
acuminate leaves which are also thinner in texture. The two upper segments of 
the corolla are united for more than half their length to form an apically bilobed 
upper lip. The only two known localities are some 500 km. apart. Details added 
from Humbert's specimen are cited in the Latin description in square brackets 
e.g. [vel pubescentes], 

II, Oplonia vincoides (Lam.) Stearn, comb, nov, (PI, 41, 46 ; Text-fig. 17A, B). 

Justicia vincoides Lam. in Encycl. Meth. Bot. i : 628 (1785). — Vahl, Enum. PI. i : 167 

(1804). 
Eranthemum vincoides (Lam.) R. Br. ex Roem. & Schult. in L., Syst. Veg., ed. nov. i : 176 

(1817). 
Anthacanthus vincoides (Lam.) Nees in DC, Prodr. 11 : 461 (1847). 

Forsythiopsis baronii Baker in Journ. Linn. Soc. Lond., Bot. 20 :2i9, t.27 (1883) " baroni ". 
Forsythiopsis australis Scott Elliot in Journ. Linn. Soc. Lond., Bot. 29 : 37 (1891). 
Forsythiopsis vincoides (Lam.) Benoist, Cat. PL Madag., Acanth. : 15 (1939). — Benoist in 

Humbert, Fl. Madag. 182 : 106, fig. 18 no. 10-13 (1967), excl. var. minor. 

Much-branched shrub 1-3 m. high ; branches slender, pilose when young ; spines 
absent. Leaves short-stalked, the blade narrowly elliptic to narrowly obovate, 
the apex usually obtuse, often emarginate, the base cuneate or attenuate, 2-9 cm. 
long, 0-6-3 cm. broad, conspicuously veined, puberulous when young, later almost 
glabrous ; petiole 2-6 mm. long. Flowers solitary or 2-3 in axillary fascicles ; 



3i6 A SURVEY OF THE TROPICAL GENERA 

pedicels 1-3 cm. long. Calyx 4-^-6 mm. long, the segments very narrowly triangular, 
sparsely puberulous. Corolla white, blue or violet, glabrous or pubescent, the tube 
c.5-8 mm. long, the segments c.8-12 mm. long, 3-5-5 mm, broad, the upper two 
free to the lower 2-5 mm. Stamens almost included or long-exserted, the filaments 
C.I -5-3 mm. or 6-8 mm. long, papillose, the anthers i -5-2 mm. long. Style puberulous. 
Capsule c.2-2-5 cm. long. 
Endemic to Madagascar. 

Madagascar : Without locality, but apparently from Fort Dauphin, Commerson 
(P, type collection oi Justicia vincoides). Fort Dauphin, Scott Elliot 2586 (BM ; K, 
lectotype of Forsythiopsis australis), 2yoi (K) ; Cloisel 2og (BM). 10 km. W. of 
Fort Dauphin, 0-12 m., 2 Sept. 1968, McWhirter & Capuron 184 (K). Central Mada- 
gascar [Imerina], Baron igjy (K, type of F. haronii). Ankorafantsika, Ambalo- 
Boeni, 10 Oct. 1950, Ramamoufisoa 2058 (P). 

Lamarck described his Justicia vincoides from herbarium material collected in 
Madagascar by Philibert Commerson, who spent the last three months of 1770 
botanizing around the French settlement. Fort Dauphin, in the extreme southeast 
of the island. Here, therefore, must be its type-locality. No specimen now exists 
in Lamarck's private herbarium and possibly he never possessed one but based his 
description of /. vincoides on the Commerson specimen still existing in the her- 
barium of the Museum National d'Histoire Naturelle, Paris (PI. 46). In 1847 
Nees referred it to the genus Anthacanthus after studying another Commerson 
specimen in Vahl's herbarium at Copenhagen. The epithet vincoides was suggested 
by a general resemblance in leaf to a periwinkle, probably Vinca minor L. The 
leaves of the Commerson specimen in Paris are elliptic, narrowly elliptic or narrowly 
obovate, cuneate at base, obtuse at the apex, with a minute apiculus, 2-4 cm. long, 
7-15 mm. broad, prominently veined and glabrous. The flowers are clustered in 
short-peduncled axillary cymes. The corolla is pubescent inside and out, the tube 
about 5 mm. long, the segments about 8 mm. long, 4 mm. broad ; the two segments 
of the upper lip are fused for the lower 2-5 cm. The filaments of the stamens are 
about 1.5 mm. long. Likewise short stamens occur on a specimen collected near 
Fort Dauphin by Cloisel {Cloisel 209). It is almost identical with the specimens 
collected near Fort Dauphin by Scott Elliot on which in 1891 he based his Forsythiop- 
sis australis ; these, however, have filaments about 5 mm. long. This Madagascan 
species thus manifests the same floral dimorphism, some individuals having long J ( 
stamens and short styles, others short stamens and long styles, as the West Indian 
species of Opionia. 

In 1883 Baker founded the genus Forsythiopsis with one species, F. haronii, on j 
specimens collected by the English missionary Richard Baron in central Madagascar, 
the precise locality in the Imerina region unfortunately not recorded. Whereas in 
the Fort Dauphin material of 0. vincoides the leaves are evidently mature, on Baron's 
specimen the leaves are only partly developed, the longest only 2-5 cm. long, 1-2 cm. 
broad, which may be associated with a local difference of climate. The corolla is 
pubescent, with the tube about 7 mm. long, the segments about lo-ii mm. long, 
4-5 mm. broad, the two segments of the upper lip being united for about the lower 



I 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 317 



2 mm . The filaments of the stamens are about 2 -5 mm . long. A specimen with similar 
immature leaves was collected by Ramamoufisoa. Benoist in 1939 equated it with 
Lamarck's Justicia vincoides and Scott Elliot's Forsythiopsis australis. Baker 
compared his Forsythiopsis with Ruellia and Benoist in 1967 similarly placed it in 
the " Tribu des Ruelliees " next to Strohilanthus. On account of its 3-colporate 
usually prolate pollen, with a pseudocolpus (colpoid streak) each side of the colpus, 
and other characters, including imbricate aestivation, it belongs, however, to the 
subfamily Ruelloideae tribe Ruellieae subtribe Odontoneminae as defined by Breme- 
kamp (1965), not to his subtribe Ruellinae or subtribe Strohilanthidinae, and it 
cannot be separated from Oplonia. The unlikelihood of Madagascan material 
having any affinity with American and the consequent neglect of comparison 
together account for the acceptance of Forsythiopsis as an endemic Madagascan 
genus. 

Although the application of the epithet vincoides to the taxon of the Fort Dauphin 
area with narrowly elliptic leaves up to 9 cm. long, 3 cm. broad, is thus certain, 
the circumscription of the species presents much difficulty. Thus Benoist in 1930 
described a. Forsythiopsis minor with obovate or narrowly obovate leaves only 2-7 mm. 
long, 1-4 mm, broad, which in 1967 he treated as F. vincoides var. minor. In 1949 he 
described under F. vincoides two varieties, one with glabrous leaves, his var. meri- 
dionalis, the other with densely hairy leaves, his var, vestita, both with obtuse 
leaves not more than 3 cm, long. He also included in F. vincoides material with ovate 
distinctly acuminate leaves. His 1967 concept of F. vincoides thus embraced several 
distinguishable taxa, which seem to represent at least three species. 

12. Oplonia puberula Steam, sp. nov. (Text-fig, 18). 

Frutex (liana caule crasso, fide Humbert et Capuron), inermis ; ramuli pallidi, 
hornotini puberuli, annotini decorticantes glabri, internodis ad 5 cm. longis. Folia 
breviter petiolata ; lamina anguste obovata, apice rotundata, obtusa vel acuta, 
basi cuneata, immatura c. 2-4*5 cm. longa, 1-2 cm. lata, supra sparse puberula, 
subter praecipue secus venas multo magis puberula, venis arcuatis conspicuis 
utroque latere costae 5-7 ; petiolus ad 6 mm. longus, dense puberulus. Flores axil- 
lares plerumque 4 ; pedicelli 5-10 mm. longi, dense puberuli. Calyx 2-5-3 mm. 
longus, omnino dense puberulus, segmentis anguste oblongis obtusis. Corolla 
violacea (fide Humbert et Capuron), bilabiata ; tubus c,7-8 mm. longus, puberulus ; 
labium superum ascendens c. 12 mm. longum, segmentis duobus fere ad medium 
coalitis quamobrem bilobatum lobis liberis c. 7 mm. longis, 3 mm. latis ; labium 
inferum descendens trifida, segmentis ellipticis obtusis c. 12 mm. longis, 5-5 mm. 
latis ; segmenta extra intraque puberula. Stamina in fauce affixa, exserta ; fila- 
menta valde papillosa, in typo 4 mm. longa ; anther ae olivaceae (fide Humbert et 
Capuron), c.i-5 mm. longae ; staminodia minuta. Ovarium glabrum ; stylus dense 
puberulus, c.12 mm. longus. 

Shrub (a thick-stemmed liane according to Humbert & Capuron) ; branches puberu- 
lous when young ; spines absent. Leaves short-stalked, the blade narrowly obovate, 
the apex rounded, obtuse or acute, the base cuneate, about 2-4-5 cm. long, 1-2 cm. 



3i8 



A SURVEY OF THE TROPICAL GENERA 



broad in an immature state, above sparsely puberulous, below much more thickly 
puberulous, particularly along the veins ; petiole to 6 mm. long. Flowers about 4 in 
the axils ; pedicels 5-10 mm. long, densely puberulous. Calyx 2-5-3 mm. long, 
densely puberulous all over, the segments narrowly oblong. Corolla violet, puberu- 
lous, the tube 7-8 mm. long, the segments c.12 mm. long, 3-5-5 mm. broad, the upper 
two united almost to the middle. Stamens exserted, the filaments (on the type) 
C.4 mm. long, the anthers c.i-5 mm. long. Style densely puberulous. 
Endemic to northern Madagascar. 




Fig. 18. Oplonia puberula Stearn. A, leaf; B, flower in front view; C, corolla in longitudinal 

section; D, calyx; E, gynoecium. 



Madagascar : Vallee inferieure de I'Androranga, affluent de la Bemarivo (nord- 
est) aux environs d'Antongondriha, Mont Anjenabe, foret ombrophile sur gneiss ; 
grande liane a tige epaisse, ecorce subereuse, corolle violette, antheres olivatres, 
style blanc ; 750 m., 3-7 Nov. 1950, Humbert S- Capuron 24103 (P, holotype ; BM). 

This taxon comes very close to 0. vincoides, from which it may be distinguished 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 319 

by its narrowly oblong calyx-segments completely and densely covered outside with 
very short erect hairs ; those of 0. vincoides are narrowly triangular with much sparser 
hairs not completely covering their surface. It is everywhere hairier than 0. vin- 
coides and has a shghtly shorter calyx and slightly larger corolla. There would also 
appear to be a difference in habit if Humbert and Capuron's description of it as a 
large liane is correct and applies to other specimens. On these grounds it is here 
given separate specific rank. The type-gathering evidently represents a long-styled 
state ; the style itself is more densely hairy and slightly longer (i.e. c.i2 mm. long 
as against 8-9 mm.) than in long-styled 0. vincoides. 

13. Oplonia minor (Benoist) Stearn, comb. nov. 

Forsythiopsis minor Benoist in Bull. Soc. Bot. France 76 : 1033 (1930). 

Forsythiopsis vincoides var. meridionalis Benoist in Mem. Soc. Bot. France 1949 : 94 (1949). 

— Benoist in Humbert, Fl. Madag. 182 : 108 (1967). 
Forsythiopsis vincoides var. vestita Benoist, loc. cit. (1949). — Benoist, loc. cit. (1967). 
Forsythiopsis vincoides var. minor (Benoist) Benoist in Humbert, Fl. Madag. 182 : 109 

(1967). 

Shrub ; branches slender, pubescent when young ; spines absent. Leaves short- 
stalked, the blade obovate or narrowly obovate, the apex obtuse, the base cuneate, 
C.2-30 mm. long, i-io mm. broad, puberulous when young, later glabrous or puberu- 
lous ; petiole c.i mm. long. Flowers solitary or paired ; pedicels 7-9 mm. long. 
Calyx 2-4 mm. long, the segments subulate. Corolla pale violet or lilac, pubescent, 
the tube c.5-6 mm. long, the segments c.8-10 mm. long, 3-4 mm. broad, the upper 
two united at base for the lower third of their length. Stamens almost included or 
long-exserted, the filaments 0.2-5 ^^i^- or 5 i^i^- ^^ng, papillose, the anthers 0.1-5 mm. 
long. 5(y/g puberulous or almost glabrous. Capsule c.i'$ cm. long. 

Var. minor 

F. minor Benoist, loc. cit. (1930). 

Leaves 2-9 mm. long, 1-4 mm, broad, puberulous when young, later glabrous. 
Calyx 2-3 mm. long, glabrous. 

Madagascar : Fort Dauphin, Behara, 10 July 1926, Decary 42^3 (BM ; K). 
Environs de Tulear, 10-200 m., Sept. 1924, Humbert & Perrier de la Bathie 2^06 (K). 
Many other localities are cited by Benoist (1967). 

Var. meridionalis (Benoist) Stearn, comb. nov. (Text-fig. 17D). 
F. vincoides var. meridonalis Benoist, loc. cit. (1949). 

Leaves 12-25 ^n^- long, 4-10 mm. broad, becoming glabrous. Calyx c.3 mm. long, 
almost glabrous. 

Madagascar : Fort Dauphin, c. 1890, Chisel 9 (BM). Amborombe, 7 Aug. 
1931, Decary gii8 (BM ; K). Massif de I'Andohahelo, au-dessus de Pisopiso 
(Beake), looo-iioo m., Jan. 1934, Humbert 13676 (K, isotype of F. vincoides 
var. meridionalis). 



320 A SURVEY OF THE TROPICAL GENERA 

Var. vestita (Benoist) Steam, comb, nov, 

F. vincoides var. vestita Benoist, loc. cit. (1949). 

Leaves 8-30 mm. long, 3-15 mm. broad, pubenilous on both surfaces, densely 
so along the veins below. Calyx 2*5-6 mm., pubenilous. 

Madagascar : Basse valleeeduFiherenana, 50-200 m., Nov. 1933, Humbert 1153^ 
(K, isotype of F. vincoides var. vestita). Foret de Besomaty entre le Fiherenana et 
risahaina (Mangoky), 750-800 m., Oct. 1933, Humbert 112^2 (P). La Table 
Tulear, Dequaire 32g (P). Au delta de la Linta, au nord d'ltampolo, i-io m., Aug. 
1928, Humbert & Swingle 5382 (P). 

The plants assembled here under the name 0. minor are primarily distinguished 
from 0. vincoides by their much smaller leaves and somewhat smaller flowers. 
Since they vary between themselves and merit study with more extensive material, 
it has seemed best to indicate variation as above, following Benoist, with some hesita- 
tion and doubt as to the validity of these taxa. 

14 Oplonia linifolia (Benoist) Stearn, comb. nov. 

Forsythiopsis linifolia Benoist in Notulae Syst. Paris 12 : ii (1945). — Benoist in Humbert, 
Fl. Madag. 182 : 109, fig. 18 no. 14 (1967). 

Shrub ; branches slender, glabrous ; spines absent. Leaves sessile, linear, the 
apex acute, the base cuneate, 5-10 mm. long, scarcely i mm. broad, glabrous. 
Flowers solitary ; pedicels 3-6 mm. long, glabrous. Calyx 2 mm. long, glabrous, the 
segments narrowly triangular. Corolla lilac, pubescent, the tube c.5 mm. long, the 
segments c.6 mm. long. Stamens slightly exserted ; filaments c.i mm. long ; 
anthers c.i mm. long. Style puberulous. Capsule c.i-5 cm. long. 

Madagascar : Bassin de la Linta, plateau calcaire a Test du delta, 200-250 m., 
August 1928, Humbert & Swingle ^^oy (P, holotype). 

This minute and rare species, only known from one locality in southern Madagascar, 
stands apart from all others by its very narrow linear leaves. 

excluded species 

Anthacanthus bispinosus Griseb. in Mem. Amer. Acad. Arts & Sci., New Ser. 8 : 525 
(1862) from Cuba (Oriente) = Barleriola solanifolia (L.) Oerst. ex Lindau in 
Engler, Bot. Jahrb. 18 : 63 (1893), based on Barleria sdanifolia L., Sp. PI. 2 : 636 
(1753) ; cf. Lindau in Urban, Symb. Antill. 2 : 204 (1900). Oersted described 
the genus Barleriola in Vidensk. Medd. Naturh. For. Kjobenh. 1854 : 136 (1855) 
but did not publish the combination Barleriola solanifolia often cited from there. 

Anthacanthus repandus (Forst. f.) Nees in DC, Prodr. 11 : 462 (1847), based on 
Justicia repanda Forst. f., Florul. Ins. Austr. Prodr. : 3 (1786), from Tanna, New 
Hebrides = Eranthemum repandum (Forst. f.) R.Br, ex Roem. & Shult. in L., 
Syst. Veg., ed. nov. i : 175 (1817) ; cf. Seem., Fl. Vit. : 185 (1866) = Pseuderanthe- 
mum repandum (Forst. f.) Guillaumin in Ann. Mus. Colon. Marseille, ser. 6, 5-6 : 49 
(1948). 



I 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 321 

Anthacanthus sinuatus Nees in DC, Prodr. ii : 462 (1847), based on Justicia sinuata 
Vahl, Symb. Bot. 2 : 11 (1791), nom. superfl. — Eranthemum longifolium (Forst. f.) 
Seem., Fl. Vit. : 185 (1866), based on Justicia longifolia Forst. f., Florul. Ins. Austr. 
Prodr. : 4 (1786), from Tanna, New Hebrides. = P sender anthemum longifolium 
(Forst. f.) Guillaumin in Bull. Mus. Hist. Nat. Paris, Ser. 2, 14 : 456 (1943). 

Anthacanthus sprengelii Nees in DC, Prodr. 11 : 461 (1847), ^ow. superfl. based on 
Eranthemum acanthodes Spreng. in L., Syst. Veg., ed. 16, i : 88 (1824), from 
Hispaniola = Scolosanthus acanthodes (Spreng.) Urban, Symb. Antill. i : 481 (1900). 
This belongs to Ruhiaceae. 

REFERENCES 

Alain, Brother (Liogier, E. E.). 1953. El endemismo en la flora de Cuba. Mem. Soc. Cub. 

Hist. Nat. 21 : 187-193. 
1958. La flora de Cuba : sus principales caracteristicas, su origen probable. Revista Soc. 

Cub. Bot. 15 : 36-59, 84-96. 
Alston, A. H. G. 1952. A revision of the West Indian species of Se/a^me//a. Bull. Brit. Mus. 

{Nat. Hist.), Bot. 1 : 25-47. 
Anderson, E. 1952. Plants, Man and Life. Boston, Mass. 
1957. ^ semigraphical method for the analysis of complex problems. Proc. Nat. Acad. 

Sci. Washington 43 : 923-927. 
Arnott, H. J. 1959. Leaf clearings. Turtox News 37 : 192-194. 
AsPREY, G. F. & RoBBiNS, R. G. 1953. The vegetation of Jamaica. Ecological Monographs 

23 : 359-412. 
Bahadur, B. 1963. Heterostylism in Oldenlandia umbellata L. Journ. Genet. 58 : 429-440. 
Benoist, R. 1967. Acanthacees. In Humbert, H., Flore de Madagascar et des Comores 182. 
Bremekamp, C. E. B. 1965. Delimitation and subdivision of the Acanthaceae. Bull. Bot. 

Surv. Ind. 7 : 21-30. 
Blanc, C. P. 1970. Analyse des facteurs de I'environment dans I'aire de repartition de Chalarodon 

madagascariensis, leurs implications biologiques. Ann. Univ. Madag., Sci. Nat. Math. 7 : 

279-306. 
Brizicky, G. K. 1969. Subgeneric and sectional names : their starting points and early 

sources. Taxon 18 : 643-660. 
BuLLARD, E., Everett, J. E. & Smith, A. G. 1965. The fit of the continents around the 

Atlantic. Phil. Trans., ser. A 258 : 41-51. 
BuRCK, W. 1883-84. Sur 1 'organisation florale chez quelques Rubiacees. Ann. Jard. Bot. 

Buitenzorg 3 : 105-119 (1883) ; 4 : 12-87 (1884). 
Camp, W. H. 1947. Distribution patterns in modern plants and the problems of ancient 

dispersals. Ecol. Monogr. 17 : 159-183. 
Cole, A. J. Ed. 1969. Numerical Taxonomy ; Proceedings of the Colloquium in Numerical 

Taxonomy held in the University of St. Andrews, September 1968. London & New York. 
CoNDE, J. A. 1952. La flora de Cuba. Mem. Soc. Cub. Hist. Nat. 21 : 123-149. 
Croizat, L. 1952. Manual of Phytogeography. The Hague. 

Darwin, C. 1877. The different Forms of Flowers on Plants of the same Species. London. 
DiETz, R. S. & Holden, J. C. 1970. The breakup of Pangaea. Sci. Amer. 223 (4) : 30-41. 
EsTABROOK, G. F. 1966. A mathematical model in graph theory for biological classification. 

Journ. Theoret. Biol. 12 : 297-310. 
Flories, G. 1970. Suggested origins of the Mozambique Channel. Trans. Geol. Soc. South Afr. 

73 : 1-16. 
Fryxell, p. a. 1969. The genus Cienfuegosia Cav. (Malvaceae). Ann. Missouri Bot. Card. 

56 : 179-250. 
Germeraad, J. H., Hopping, C. A. & Muller, J. 1968. Palynology of Tertiary sediments from 

tropical areas. Rev. Palaeobot. &■ Palyn. 6 : 189-348. 



322 A SURVEY OF THE TROPICAL GENERA 

GowER, J. C. & Ross, G. J. S. 1969. Minimum spanning trees and single linkage cluste 

analysis. Applied Statistics 18 : 54-64. 
Green, P. S. 1969. Studies in the genus Jasminum L. : IV. The so-called New World] 

species. Kew Bull. 23 : 273-275. 
Grisebach, a. H. R. 1865. Die geographische Verbreitung der Pflanzen Westindiens. 

Ahhandl. Kon. Ges. Wiss. Gottingen 12, Phys. : 3-80. (Reprinted in Grisebach, Gesammelte 

Abhandlungen : 222-285 (1880) ). 
Hatheway, W. H. 1962. A weighted hybrid index. Evolution 16 : i-io. 
Hawkes, J. G. & Smith, P. 1965. Continental drift and the age of angiosperm genera. 

Nature, Lond. 207 : 48-50. 
Hedge, I. C. 1961. Some remarks on the perennial species of Ziziphora with special reference 

to those in Turkey. Notes R. Bot. Gard. Edinb. 23 : 209-221. 
Hepper, F. N. 1965. Preliminary account of the phytogeographical afi&nities of the flora of 

West Tropical Africa. Webbia 19 : 593-617. 
Irwin, H. S. & Rogers, D. J. 1967. Monographic studies in Cassia (Leguminosae-Caesal- 

pinioideae). II : A taximetric study of section Apoucouita. Mem. New York Bot. Gard. 

16 : 71-120. 
Johnson, L. A. S. 1968. Rainbow's end : the quest for an optimal taxonomy. Proc. Linn. 

Sac. New South Wales 93 : 8-45. 
Lawralree, a. 1943. Les genres Hoffmaniella Schlecht., Eleutheranthera Poit. et Exomio 

carpon Lawralree gen. nov. Bull. Jard. Bot. Brux. 17 : 55-64. 
Lehmann, H. 1954. Der tropische Kegelkarst auf den Grossen Antillen. Erdkunde 8 

130-139. 
Leonard, E. C. 1951-58. The Acanthaceae of Colombia. Contrib. U.S. Nat. Herb. 31. 
LiNDAU, G. 1893. Beitrage zur Systematik der Acanthaceen. Bot. Jahrb. 18 : 36-64. 

1895. Acanthaceae. In Engler & Prantl, Nat. Pflanzenfam. 4 (36) : 274-354. 

1900. Acanthaceae. In Urban, Symb. Antill. 2 : 170-250. 

Melville, R. 1966. Continental drift, Mesozoic continents and the migrations of the angio 

sperms. Nature, Lond. 211 : 1 16-120. 
MuLLER, J. 1970. Palynological evidence on early differentiation of angiosperms. Biol. Rev. 45 

417-450. 
Nees von Esenbeck, C. G. 1847. Anthacanthus. In De CandoUe, Prodromus Systemati 

Naturalis Regni Vegetabilis 11 : 460-462. 
Perrier de la Bathie, H. 1936. Biogeographie des Plantes de Madagascar. Paris. 
Prance, G. T., Rogers, D. J. & White, F. 1969. A taximetric study of an angiospe 

family : generic delimitation in the Chrysobalanaceae. New Phytol. 68 : 1 203-1 234. 
ScHOPF, J. M. 1970. Relation of floras of the southern hemisphere to continental drift 

Taxon 19 : 657-674. 
ScHUCHERT, C. 1935. Historical Geology of the Antillean-Caribbean Region. New York. 
Seifriz, W. 1943. The plant life of Cuba. Ecol. Monogr. 13 : 375-426. 
Sims, R. W. 1966. The classification of the Megascolecoid earthworms ; an investigation o; 

Oligochaete systematics by computer techniques. Proc. Linn. Soc. Lond. 177 : 125-141. 
Sleumer, H. 1938. Die Gattung Agauria (DC.) Hook. f. Bot. Jahrb. 69 : 374-394. 
Smith, A. G. & Hallam, H. 1970. The fit of the southern continents. Nature, Lond. 225 

139-144. 
Sneath, p. H. a. 1962. The construction of taxonomic groups. Symp. Soc. Gen. Microbiol 

12 : 289-332. 
Stearn, W. T. 1959. A botanist's random impressions of Jamaica. Proc. Linn. Soc. Lond, 

170 : 134-147- 
1965. Grisebach's Flora of the British West Indian Islands : a biographical and bibli 

graphical introduction. Journ. Arnold Arb. 46 : 243-285. 
1969. The Jamaican species of Columnea and A lloplectus (Gesneriaceae) . Bull. Brit. Mus. 

{Nat. Hist.), Bot. 4 : 179-236. 



OPLONIA AND PSILANTHELE (ACANTHACEAE) 323 

Stevens, P. F. 1970. Agauria and Agarista : an example of tropical transatlantic affinity. 

Notes R. Bot. Gard. Edinh. 30 : 341-359. 
Thomas, J. L. i960. A monographic study of the Cyxillaceae. Contrib. Gray Herb. 

Harvard Univ. 186. 
1 96 1. Schizocardia belizensis : a species of Purdiaea (Cyxillaceae) from Central America. 

Journ. Arnold Arb. 42 : iio-iii. 
Underwood, L. M. 1905. A summary of Charles Wright's explorations in Cuba. Bull. Torrey 

Bot. Club 32 : 291-300, map. 
Urban, I. 1902. Notae biographicae peregrinatorum Indiae occidentalis botanicorum. 

Symb. Antill. 3 : 14-158. 
Vester, H. 1940. Die Areale und Arealtypen der Angiospermen-Famihen. Bot. Arch. 41 : 

203-275. 295-356, 520-577. 
VuiLLEUMiER, B. S. 1967. The origin and evolutionary development of heterostyly in the 

angiosperms. Evolution 21 : 210-226. 
Weyl, R. 1965. Die palaogeographische Entwicklung des mittelamerikanisch-westindischen 

Raumes. Geol. Rundschau 54 : 1213-1240. [English translation in Bol. Inform. Asoc. 

Venezol. Geol. 9 : 99-120 (1966) ]. 
Wernham, H. F. 1914. A Monograph of the Genus Sabicea. London. 
Wilkinson, C. 1967. A taxonomic revision of the genus Teldenia Moore (Lepidoptera : 

Drepanidae, Drepaninae). Trans. R. Entom. Soc. London 119 : 303-362. 
WiRTH, M., Estabrook, G. F. & Rogers, D. J. 1966. A graph theory model for systematic 

biology, with an example for the Oncidiinae (Orchidaceae) . Syst. Zool. 15 : 59-69. 



iV 



William T. Stearn, D.Sc, Sc.D. 

Department of Botany 

British Museum (Natural History) 

Cromwell Road 

London, S.W.7 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 38 




Pollen grain of Oplonia arniata var. pallidior Stearn [Steam y26) 
X 2000 ; below, polar view, x 2000. 



above, equatorial view, 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 39 





Pollen grain of Oplonia jamaicensis (Lindau) Stearn (Proctor 8423) ; above, equatorial 
view, X 2000 ; below, surface, x 6500. 



lUill. Br. Mus nat. Hist. (Bot.) 4, 7 



PLATE 40 





(a) Pollen grain of Psilanthele eggersii Lindau {Eggers I5i2g) ; equatorial view, x 2850. 

(b) Pollen grain of Oplonia grandiflora (Lindau) Stearn {Ule 6493) ; polar view, x 3000. 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 41 




Pollen grains of Oplonia vincoides (Lam.) Stearn {McWhirter &■ Capuron 184) ; above, 

X 1000 : below, polar view, x 2000. 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 42 




Oplonia grandiflora (Lindau) Stearn ; isotype {Ule 64g3). 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 



43 



(a) 



(b) 









/ ' 


1 














>v 


. >^^ 




-^v „. . 


. 


>, 




-6 


"i^' 




> 


*X^M[^*' 


— -- 


''~- s^ 


■ '^ 


s 




.•)»' 


m ■ 



•o- 



■■^ 



t^«- 





(a) Oplonia spinosa (Jacq.) Raf. ; drawing from Jacquin, Select. Stirp. Anier. Pict. (1780) 

representing /ws^zcza s^inosa Jacq. ; x \. 

(b) Cleared leaves of Oplonia armata var. armata {Proctor ggsj) to illustrate veining 

characteristic of genus Oplonia ; X 3. 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 44 




Oplonia microphylla (Lam.) Stearn ; isotype (Richard). 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 



45 




Oplonia nannophylla (Urban) Stearn ; holotype [Ekman 15634). 



U. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 46 




' ' 1! ' '' '2! f! 

British Muskum (Natvral HistoryV 



l| 2! 3 4| 5! 

'"liinillllliiilHi iiihiiihmliuihml 



i- 



HtRBARILM MTStl BRIIANNICI 



Oplonia vincoides (Lam.) Steam; syntype (Madagascar, Commerson) 



\ 



Bull. Br. Mus. nat. Hist. (Bot.) 4, 7 



PLATE 47 




lll l i 1 1 r [ 1 1 1 M |n MM It 

BunifH MuisvM (NAnmAl Hnroav) 
pHOTtV.RHPH 



3 J 41 1] 

lllllllM'llll'lllllllll'llllI 






lIlRBARIfM .MI.SEI BRITANNICJ 






Psilanthele eggersii Lindau ; isotype {Eggers I5i2g). 



\ 



p^' 



?.'■?■«::- 










BULLETIN OF THE 
BRITISH MUSEUM (NATURAL HISTORY): 



BOTANY 
VOLUME 4 

1. Cuticular studies as an aid to taxonomy. By C. A. Stage. 1965. £1.75 

2. The genus El ,phoglossum in the Indian Peninsula and Ceylon. By 

W. A. Sledge. 1967. £o-35 

3. Fungi of recent Nepal expeditions. By F. L. Balfour-Browne. 

1968. i^-9^ 

4. A synopsis of Jamaican Myrsinaceae. By W. T. Stearn. 1969. ;£i.20 

5. The Jamaican species of Columnea and Alloplectus {Gesneriaceae) . 

By W. T, Stearn. 1969. £140 

6. New or little known Himalayan species of Swertia and Veratrilla 

{Gentianaceae). By H. Smith. 1970. £i-55 

7. A survey of the tropical genera Oplonia and Psilanthele 

[Acanthaceae). By W. T. Stearn. 1971. £2.60 



PRINTED IN GREAT BRITAIN BY ALDEN & MOWBRAY LTD AT THE ALDEN PRESS, OXFORD 



ANGIOSPERMS OF THE ISLANC)#^ 
OF THE GULF OF GUINEA ^ 

(FERNANDO PO, PRINCIPE; 
S. TOME, AND ANNOBON) 



A. W. EXELL 




BULLETIN OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 
BOTANY Vol. 4 No. 8 

LONDON: 1973 



ANGIOSPERMS OF THE ISLANDS OF THR 

GULF OF GUINEA (FERNANDO PO, 

PRINCIPE, S. TOME, AND ANNOBON) 




BY 

ARTHUR WALLIS EXELL 



K 



Pp 325-411 



BULLETIN OF 

THE BRITISH MUSEUM (NATURAL HISTORY) 

BOTANY Vol. 4 No. 8 

LONDON: 1973 



THE BULLETIN OF THE BRITISH MUSEUM 

(natural history), instituted in 1949, is 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they 
become ready. Volumes will contain about three or 
four hundred pages, and will not necessarily be 
completed within one calendar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 
Department. 

This paper is Vol. 4, No. 8 of the Botanical series. 



World List abbreviation : 
Bull. Br. Mus. nat. Hist. (Bot.) 



Trustees of the British Museum (Natural History), 1973 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY] 



Issued 14 December 1973 



Price £3.45 



r 



ANGIOSPERMS OF THE ISLANDS OF THE 

GULF OF GUINEA (FERNANDO PO, 

PRINCIPE, S. TOME, AND ANNOBON) 

By ARTHUR W. EXELL 

The time has come when it seems useful to give an up-to-date list of the flowering 
plants of the Guinea Islands. There have been many supplements etc. since the 
publication in 1944 of the Catalogue of the Vascular Plants of S. Tome {with Principe 
and Annobon) (i)/ many changes recorded in the second edition of the Flora of 
West Tropical Africa (10, 12, 17, 18, 21), various publications by Portuguese authors 
(8, 9, 13), by Emilio Guinea (2, 3, 4, 5, 6, 7) and Antonio Escarre (19), as well as 
changes proposed by specialists in certain families or genera. Furthermore, there 
has been considerable additional collecting. 

Fernando Po has now been added because the only way to get a clear picture of 
the flora of that island is to extricate the information from F.W.T.A. ed. 2 (10, 12, 
17, 18, 21) and, as will be seen from the present work, this is at times incomplete. 

FLORA OF FERNANDO PO 

A clear distinction must be made between Fernando Po, only 32 km from the 
mainland, and the other three islands. The latter (Principe, S. Tome and Anno- 
bon) can scarcely be classified as 'oceanic islands' and as I know of no appropriate 
term in use I shall call them 'maritime islands' and Fernando Po a 'littoral island'. 
Of the three maritime islands only Annobon, the smallest and the furthest from the 
mainland, has some of the characteristics of an oceanic island, especially in its 
high proportion of Pteridophyta and of widespread species of Angiosperms. 

The flora of Fernando Po is unlikely to have been much affected by the narrow 
strip of water separating it from the continent. Fairly convincing proof of this will 
be given later. I have noticed, however, in Africa that even a short stretch of water 
may prove a barrier to many Leguminosae (except for littoral species) . I attribute 
this to the comparatively heavy seeds unsuitable for wind distribution and not, as 
far as I know, much distributed by birds. On the three maritime islands, although 
the list of Leguminosae may seem extensive, if all the introduced species, the lit- 
toral species carried by ocean currents and the doubtfully indigenous cosmopolitan, 
pan-tropical or American weeds are excluded, the family is poorly represented com- 
pared with the adjacent mainland. Although Fernando Po has more Leguminosae, 
its flora still seems impoverished with respect to this family. 

The Compositae, excellently distributed by wind, are quite well represented in 
Fernando Po but poorly so on the maritime islands when only species with some 
claim to be indigenous are considered. This apparent discrepancy is almost cer- 
tainly due to the presence of montane or submontane grassland (Moka, etc.) in 
Fernando Po and its almost complete absence in the maritime islands. 

* For references, see Bibliography, pp. 333-336. 



re 



328 ANGIOSPERMS OF THE ISLANDS 

Rubiaceae, Euphorbiaceae and Orchidaceae are well represented both in Fer- 
nando Po and in the maritime islands. 

Endemism in Fernando Po, as befits a littoral island, is much lower than in the 
maritime islands. For the relevant figures and the relationship between the islands 
see the Statistical Summary below. 

As usual in floristic comparisons one can seldom isolate the statistical behaviour 
due to separate factors but in comparing these islands at least differences due to 
climate are probably minimal, though Annobon is rather drier than the other three. 

INTRODUCED PLANTS 

All introduced plants of which I have seen records or specimens have been in- 
cluded. Species are marked with an asterisk in the Systematic List whenever there 
is a high probability that they are not indigenous. There are of course very many 
other species (mainly weeds) where the evidence is not conclusive. One reason 
for including the introduced species in the list is to facilitate the naming of future 
collections as a glance at the list may save a lot of trouble. 

The introduced species fall into the following main categories : 

(i) The successful introductions grown as crops for export, such as cocoa and coffee. 

(2) Unsuccessful introductions for the same purpose. In the early days of 
Portuguese colonization there are records of many attempted introductions (wheat,, 
rye, etc.) at a period when there was no experience of tropical agriculture to act as 
a guide. Many such species probably no longer exist on the islands. 

(3) Species introduced as shade plants for cocoa or as ground cover. 

(4) Species grown as local food plants (bananas, breadfruit, pineapples, etc.). 

(5) Ornamental plants. 

(6) Accidentally introduced tropical American weeds (usually clearly not in- 
digenous). 

(7) A few European or cosmopolitan weeds (such as Plantago major) very improb- 
ably indigenous. 

There is of course another large category of weeds and widespread species which 
are probably introduced but could be indigenous. 

It will be noticed that there are relatively far more introduced species for S. Tome 
(and to a lesser extent for Principe) than for the other two islands. The reasons 
for this are historical. Both S. Tome and Principe have been highly cultivated for 
about four centuries. By 1550 two-thirds of S. Tome were said to be under cul- 
tivation and these islands were probably the first places where tropical agriculture 
for export was carried on. Fernando Po was much less exploited and I do not know 
of any serious export from Annobon. Nor has this small island many introduced 
ornamental plants. 

STATISTICAL SUMMARY 

I have been interested for over forty years in the problem of comparing floras of 
similar or different sizes inhabiting areas of similar or different sizes in order to 
obtain Quotients of Affinity and Quotients of Similarity (see especially Exell and 



i 



OF THE GULF OF GUINEA 329 

Wild in Kirkia 2 : 108-130 (1961)). In the work cited we discussed in some detail 
the various quotients wjiich have been employed at different times and the justifi- 
cation for the use of the simple formula referred to as QA (EW) which is the ex- 
pression of the number of species common to two floras as a percentage of the 
smaller one. This is used in the present work and is here simply called QA (Quotient 
of Affinity). After further thoughts about this formula (since 1961) I have now 
the impression that it tends to become less meaningful when a very small flora is 
compared with a very large one. In the ultimate case, when the flora of an island 
or region is compared with the flora of the whole of the rest of the world we get a 
QA which is exactly the inverse of the proportion of endemism in the small flora 
(i.e. with no endemics the QA becomes 100). Thus when there is a very great 
disparity of size between the floras under consideration the relative abundance or 
absence of endemism begins to loom very large in the QA figures and to obscure 
other factors. 

Where the QA employed is at its most useful and most accurate is in comparing 
the flora of a given region with those of neighbouring regions not differing greatly 
from it or from each other in size. I have borne this in mind in the statistical 
comparison of the floras of the four islands with the mainland flora and confined 
this comparison to the Southern Nigeria and West Cameroon regions, and a com- 
bination of the two (which already begins to get too large). To extend the com- 
parison to the whole tropical West African flora would be to approach the situation 
I have indicated above where the figures would do little beyond reflecting the differ- 
ent degrees of endemism on the different islands and be merely an immensely 
laborious way of arriving at such a result. 

The Quotients of Affinity of the islands with one another and with the mainland 
regions are set out in Table I. It should be stated that the figures given in this 



FP 

P 

ST 





Table 


I 








Quotients of Affinity of the flora of the islands of the 
Gulf of Guinea and the adjacent mainland 




FP P 


ST 


A 


SN 


WC 


SN + ^ 


— 59-0 
59-0 — 
47-8 67-9 


47-8 
67-9 


56-25 

44-7 

71-5 


68-5 
73-2 
64-4 


71-5 
6o-2 
52-6 


89-0 
78-9 
71-9 



56-25 44-7 71-5 — 72-1 55-8 77-9 



FP, Fernando Fo, 1105 species; P. Principe. 314 species; ST, S. Tome, 601 species; A, Annobon, 208 
species; SN, Southern Nigeria, 'S. Nig.' of the Flora of West Tropical Africa; WC, West Cameroon, 
'Brit. Cam.' '[Br.] Cam.' and 'W. Cam.' of the various volumes of the Flora of West Tropical Africa. 

table are not absolutely correct because of slight adjustments made to the sys- 
tematic list right up to the time of going to press. The mathematical discrepancy 
is, however, negligible, especially in view of the considerable and unavoidable 
subjective factor involved in choosing which taxa to exclude (such as possibly 
introduced or too doubtfully identified species). 

If the QA figures between the various islands are compared with similar statistics 
given in the past (see especially Exell and Wild, loc. cit. Table I, column QA (EW)) 



330 ANGIOSPERMS OF THE ISLANDS 

it will be seen that relationships between Principe, S, Tome and Annobon have 
not altered appreciably although the total number of species recorded for each 
island has increased. The figures for Fernando Po have, however, changed sig- 
nificantly. This is mainly due to a complete re-appraisal of the Fernando Po 
flora in the second edition of the Flora of West Tropical Africa. This has resulted 
in a great increase in the QA's between Fernando Po and the three maritime islands 
correlated with a marked decrease in the percentage of species considered endemic 
to Fernando Po (from 12 per cent in 1944 to 3-6 per cent now) ; this is due partly 
to taxonomic changes (reduction of endemics to synonymy), partly to the dis- 
covery on the mainland of species formerly thought to be endemic and partly to an 
increase of non-endemic species recorded. The same process has occurred to a 
lesser extent in the other three islands (fairly significantly in Annobon) as Table II 
shows. d 

Table II 

Percentage of species considered as endemic in the ^L 

flora of the islands of the Gulf of Guinea W 





1944 


I97I 


Fernando Po 


I2*0 


3-6 (3-8) 


Principe 


12-7 


9-9 (13-7) 


S. Tome 


19-4 


15.4 (17-8) 


Annobon 


14-8 


77(11) 



The figures in brackets in the last column are those obtained if species occurring on more than one 
island but not elsewhere are reckoned as endemics. 

The decrease in Annobon is due almost entirely to the fact that the recorded non- 
endemic flora has nearly doubled since 1944. 

Relatively few new endemic species have been described to counterbalance the 
tendencies mentioned above. There is little hope, however, that the present 
figures represent anything like finahty. There are many more discoveries still to 
make. 

In the past there have not been sufficiently accurate data available to work out 
the QA's between the various islands and selected regions of the mainland. This 
has now been largely remedied with regard to Southern Nigeria and West Cameroon 
by the publication of the second edition of the Flora of West Tropical Africa. 

It is perhaps worth mentioning that the statistics were first worked out sepa- 
rately for the Dicotyledons and Monocotyledons but no significant difference was 
found. The Monocotyledons do not behave statistically as a separate unit. The 
Orchidaceae, for example, with many endemics, behave much as the Melastomata- 
ceae, Rubiaceae and Euphorbiaceae while the Gramineae and Cyperaceae behave 
very like the Leguminosae and the Compositae. 

It is now easiest to consider each of the four islands separately. 

Fernando Po 

The QA of Fernando Po with the adjacent coastal region as a whole (SN + WC 
of Table I) is 89. This is very high indeed. For comparison, the highest QA 



•I 



|L OF THE GULF OF GUINEA 331 

|fcbtained in a statistical analysis of a sample of the flora of the region covered by 
w^^lora Zamhesiaca was %y between the Eastern Division of Rhodesia and the southern 
H part of Mozambique, two adjacent and floristically closely related regions on the 
j mainland of Africa. The only QA higher than 89 that I have yet recorded is 92-4 
I between the floras of Guernsey and Alderney (Channel Islands). This QA of 89 
j between Fernando Po and SN + WC gives statistical proof of what has long been 
recognized, that Fernando Po does not have an insular flora but is floristically part 
of the mainland. 

The QA 71-5 of Fernando Po with West Cameroon is slightly higher than the 
QA 68-5 between the island and Southern Nigeria, indicating a fairly equal deriv- 
ation of the flora of Fernando Po from West Cameroon and Southern Nigeria, 
but, if the species contributing to these two QAs are studied in detail, the impression 
is gained that the slight difference in favour of West Cameroon should really be 
greater than it is for it is certainly based on more significant distribution patterns. 
Among the species common to Fernando Po and Southern Nigeria there are far 
more widespread species and weeds, and many of these probably occur also in West 
Cameroon, which appears to be somewhat under-collected. 

Principe 

The QA of this island's flora with that of the adjacent coastal region as a whole 
(SN + WC) is 78-9, showing a high affinity between the two, but clearly indicating 
that the flora of Principe is more insular than that of Fernando Po. The QA with 
Southern Nigeria is 73-2 and that with West Cameroon 60-2, very considerably 
different from the figures for Fernando Po, for they show that Southern Nigeria 
is a more important source than West Cameroon for the flora of Principe. The 
QA of 59*0 with Fernando Po and 67-9 with S. Tome are both fairly high, as com- 
pared for example with a QA of 47-8 between Fernando Po and S. Tome, and seem 
to indicate some derivation of the flora of Principe from the two larger islands 
between which it lies. 

5. Tome 

The QA between this island and the adjacent coastal region as a whole (SN + WC) 
is 71-9, appreciably lower than the islands nearer the mainlaind and probably fairly 
typical of an insular flora with considerable endemism. The QA 64-4 with Southern 
Nigeria is considerably higher than the QA 52-6 with West Cameroon, indicating 
that Southern Nigeria is the more important as a source of the flora of S. Tome. 

Annobon 

This, although further from the mainland than S. Tome, has a higher QA, 77-9, 
with the adjacent coastal region as a whole (SN + WC). Here we begin to get, I 
believe, a partial 'oceanic island' effect, a somewhat impoverished flora with a 
higher proportion of widespread species and rather fewer endemics. The affinity 
with Southern Nigeria (QA 72-1) is much higher than that with West Cameroon 
(QA55-8). 



332 ANGIOSPERMS OF THE ISLANDS 

A Special feature of the Annobon figures is the high QA, 71-5, with S. Tome, and 
this appears to be a real affinity, indicating that the small Annobon flora has been 
partly derived from the much larger and comparatively near-by island of S. Tome. 

The QA of 56-25 between Annobon and Fernando Po seems abnormally high for 
that between the two most widely separated islands, especially when it is compared 
with QA 47-8 between S. Tome and Fernando Po. Detailed study indicates that is 
partly an 'oceanic island' effect, the flora of Fernando Po being more than five 
times the size of that of Annobon, and partly the result of a few anomalous dis- 
tributions, which may not all be genuine. 

General 

There has been no change in the general hypothesis that in the main each island 
received its flora independently from the adjacent regions of the mainland, each 
receiving a considerably different selection so that the QAs with the mainland are 
usually higher than those between one island and another. Each island mostly 
produced its own endemics. There are relatively few cases of species endemic to 
more than one island and some of these may not survive future investigation and 
further collecting (i.e. they may yet be found on the mainland). A few, especially 
those common to S. Tome and Annobon, are probably genuine. 

I hope that it will be apparent that the comparisons given above, and in fact 
most of the discussion, would be almost impossible without using the QA concept. 
Whether the QA used here is the best one available is not so certain and any sug- 
gested improvement will always merit examination. 



SYSTEMATIC LIST 

This list is mainly a compilation of published records of taxa from the islands but 
I have been fortunate in getting assistance, for which I am most grateful, from 
specialists in a number of famihes and from other colleagues. Acknowledgements 
are made after the bibliography. The list is very condensed in form but it is possible 
from the bibliographical references and some short notes appended under the species 
to obtain fairly full nomenclatural data and records of specimens. 

An asterisk * indicates an introduced species. It is often impossible to be certain 
whether a taxon such as a widespread weed is indigenous or not. It is fairly certain 
that all the species marked * are introduced but it is by no means certain that all 
the species not so marked are truly indigenous. 

The bibliographical references are given by number in the central column. To 
avoid unnecessary complication, where infraspecific categories are now recognized, 
the older bibliographical references have been placed where they almost certainly 
belong. This usually consists of putting the reference under the typical subspecies 
or variety when only the specific name was originally given in the work cited. The 
same has been done in a few cases where it seems quite clear to which infraspecific 
taxon (other than the typical one) the work cited must refer even when this was 
not mentioned or had not been established at the time. 



OF THE GULF OF GUINEA 333 

The letters FP, P, ST, A in the right-hand column indicate Fernando Po, Prin- 
cipe, S. Tome and Annobon respectively and the letter E in the right-hand margin 
indicates an endemic taxon. 

In the case of Fernando Po (FP) I have indicated all records additional to or 
omitted from the second edition of the Flora of West Tropical Africa (except in 
the case of introductions). These fall into very different categories. Some are 
evident accidental omissions (when, for example, the type is from Fernando Po), 
some are new and probably correct records, others are very doubtful indeed 
(especially those to which I have given question marks). 

ANNOTATED BIBLIOGRAPHY OF WORKS PUBLISHED 
FROM 1944 ONWARDS 

References to works prior to 1944 will mainly be found in i, 10, 11, 12, 17, 18, 21. 

Different numbers are given to the separate volumes (or parts) of the second 
edition of the Flora of West Tropical Africa in order to maintain chronological 
sequence and to facilitate reference. 

1. A. W. Exell. Catalogue of the Vascular Plants of S. Tome [with Principe 
and Annobon). 1944. 

References to works prior to 1944, species references and fairly full synonymies. 
Contributors (Angiosperms) other than the author were : Leguminosae (partly 
by E. G. Baker and D. Hillcoat) ; Rubiaceae (by G. Taylor) ; Apocynaceae (by 
W. R. Philipson) ; Asclepiadaceae (partly by C. Norman) ; Loganiaceae (by W. R. 
Philipson) ; Acanthaceae (partly by E. Milne-Redhead) ; Labiatae (by G. Ta^dor) ; 
Euphorbiaceae (partly by J, Hutchinson) ; Ulmaceae (by G. Taylor) ; Moraceae 
(by G. Taylor) ; Urticaceae (by G. Taylor) ; Cannabaceae (by G. Taylor) ; Mono- 
cotyledones (by J. E. Dandy, with V. S. Summerhayes assisting with Orchidaceae) . 

2. E. Guinea Lopez. Ensayo Geobotdnico de la Guinea Continental Espanola. 
1946. 

The relevant part is the 'Catalogo Floristico', pp. 218-368, in which Fernando 
Po records are given and some references made to the other islands. 

3. E. Guinea (Lopez). Agrostigrafia fernandina e hispano-guineanae. An. Agr. 
Territ. Espan. del Golfo de Guinea 1947, publ. 13 : 95-164, 1948. 

4. E. Guinea (Lopez). Juncaceae, Cyperaceaeque Fernandopoinae. An. Jard. 
Bot. Madrid 9 : 335-360. 1950. 

5. E. Guinea (Lopez). Melastomataceae Fernandopoanae. Bol. Real Soc. Esp. 
Hist. Nat. 49 : 175-187. 1951. 

6. E. Guinea (Lopez). Asteraceae (Compositae) Fernandopoanae. An. Inst. 
Bot. Cavan. 1 : 301-340. 1951. 

7. E. Guinea Lopez. Esquema de la Vegetacion de Fernando Poo. Confer- 
encia Internacional de Africanistas Occidentales (1951). Circular Paper. 
1951. 



334 ANGIOSPERMS OF THE ISLANDS 

This includes some lists of plants collected by Guinea in Fernando Po. The 
work is repeated in Spanish (41-45), French (52-56), English (62-66) and Portu- 
guese {73-77). The Araliaceae {Polyscias fulva, Schefflera mannii, S. ledermannii) 
are omitted from the English version. y 

8. L. G. Sobrinho. Vascular Plants from S. Tome. Port. Act. Biol. (B), 3 : 392- 

401- 1953- 
The date on the title-page of the separate is 1952 but it was apparently published 
in 1953 (see p. 392). This work records plants collected in Principe and S. Tome 
by d'Oliveira and Noronha in 1951. 

9. L. G. Sobrinho. Vascular Plants from Ano Bom and Principe Islands. Port. 
Act. Biol. (B) 4 : 177-190. 1953. 

This records plants (now at LISU) collected by F. Newton in Annobon in 1893 
and a few specimens collected by de Souza Jr in Principe in 1880. The existence of 
these specimens was unknown to me at the time of the pubhcation of (i). There are 
many new records from Annobon. 

10. J. Hutchinson and J. M. Dalziel. Flora of West Tropical Africa i (i), ed. 2, 
revised by R. W. J. Keay. 1954. 

Authors of various families other than Keay are indicated in the work. I am 
much indebted to this excellent second edition and the Fernando Po records have 
been largely extracted from this and the succeeding volumes. 

11. A. W. Exell. Supplement to the Catalogue of the Vascular Plants ofS. Tome (with 
Principe and Annobon). 1956. 

This brought (i) up to date by including the records in (8) and (9) and a consider- 
able collection made in S. Tome by Espirito Santo. Acknowledgement to col- 
laborators is made in the introduction. This opportunity is taken to correct two 
formulae on p. 4 in the discussion on statistics. These should read : (i) the increase 
in number of species by adding area B to area A = T — a = a logg [(A + B)/A] 
and (2) similarly the increase of species by adding area A to area B = T — 6 = 
a loge [(A + B)/B]. 

12. As (10). Vol. I (2). 1958. 

13. H. Lains e Silva. Sao Tome e Principe e a Cultura do Cafe. 1958. 

This important work has much information about the vegetation of the two 
islands, the ecology and the introduced species. It should be noted that the ' Nomes 
latinos das especies vegetais' on p. 405 does not refer only to species occurring in 
the islands but includes a few species from other parts of the world mentioned in 
the text. There are a number of new records. I have not myself seen the specimens 
of these. 

14. A. W. Exell and A, Rozeira. Aditamento a Flora das Ilhas de S. Tome e do 
Principe. 6 Conferencia Internacional dos Africanistas Ocidentais (Ilha de S. 
Tome, 1956) 3 : 77-92. 1958. 



OF THE GULF OF GUINEA 335 

Spermatophyta by A. W. Exell. This mainly consists of new records from the 
collections of A. Rozeira in Principe and S. Tome, I was not shown any proofs of 
this publication and regret a number of typographical errors for which I was not 
responsible. 

15. A. W. Exell. Additions to the Flora of S. Tome and Principe. Bull. I. F.A.N. 
21, ser. A, 2 : 439-476. i959- 

Mainly an account of new records from collections by Th. Monod made in Principe 
and S. Tome in 1956. 

16. A. W. Exell. Angiosperms of the Cambridge Annobon Island Expedition. 
Bull. Br. Mus. nat. Hist. {Bot.), 3 : 93-118 (1963). 

This listed new and interesting plants collected by F. A. Melville and T. C. Wrigley 
in Annobon in 1959. Certain families not by the author were contributed as follows : 
Loganiaceae (by J. Lewis) ; Solanaceae (by W. T. Stearn) ; Piperaceae (by W. T. 
Steam) ; Loranthaceae (by W. T. Stearn) ; Orchidaceae (by V. S. Summerhayes) ; 
Gramineae (by W. D. Clayton). 

17. As (10) but revised by F. N. Hepper. Vol. 2. 1963. 

18. As (17). Vol. 3 (i). 1968. 

19. A. Escarre. Ap'ortaciones al Conocimento de la Flora de Fernando Poo. I — 
Araliaceae, Umbelliferae. 1968. II — Piperaceae, Urticaceae. 1969. 

20. W. W. Sanford. The Orchid Flora of Equatorial Guinea in relation to that of 
West Africa. Mitt. Bot. StSamml. Miinch. 10 : 287-298. 1971. 

21. As (18). Vol. 3 (2). 1972. 

22. J. H. P. de Barros Ferreira. Contribui9ao para o Estudo das Malvaceas de 
S. Tome e Principe. Garcia de Orta, 11 : 27-55. 1963. 

23. J. H. P. de Barros Ferreira. Contribui^ao para o Estudo das Begoniaceas de 
S. Tome e Principe. Garcia de Orta, 13 : 525-543. 1965. 

24. J. H. P. de Barros Ferreira. Contribuigao para o Estudo das Melastomataceae 
de S. Tome e Principe. Garcia de Orta, 16 : 69-86. 1968. 

Includes Tristemma thomensis Barros Ferreira published by the author on p. 66 
of the same volume. 

25. J. do Espirito Santo. Nomes crioulos e vernaculos de algumas plantas de 
S. Tome e Principe. Bolm cult. Guine port. 24 : 193-211. 1969. 

In spite of the title the plants recorded are all from S. Tome. The author states 
in a footnote to the first page of (27) that the 'prospec9ao botanica' of 1968 did not 
get to Principe. Numerous spelling mistakes, etc. are mainly corrected in the 
Corrigenda to (27). 

26. J. do Espirito Santo. Algumas Plantas medicinais e venenosas de S. Tome e 
Principe. Bolm cult. Guine port. 24 : 917-940. 1969. 

Notes to (25) apply equally to this. 



336 



ANGIOSPERMS OF THE ISLANDS 



27. J. do Espfrito Santo. Aditamento a Flora de S. Tome. Bolm cult. Guine 
port. 25 : 489-511. 1970. 
New records are mainly introduced species but there are a few interesting in- 
digenous ones. Erythroxylum emarginatum, Ilex mitis and Alangium chinense add 
new families to those recorded from S. Tome. The specimens are at LISJC and 
most of them also at LISC. This work corrects many mistakes in (25) and (26) but 
adds a few more most of which are fairly obvious. 



ACKNOWLEDGEMENTS 



S. Balle : Loranthaceae. 

J. P. M. Brenan : various. 

J. F. M. Cannon : Araliaceae, Umbelliferae. 

W. D. Clayton : Gramineae. 

J. E. Dandy : various. 

A. Escarr6 : Piperaceae, Urticaceae and 

various. 
A. & R. Fernandas : Barringtoniaceae, Mela- 

stomataceae, Onagraceae. 
F. N. Hepper : various. 
C. Jeffrey : Compositae, Cucurbitaceae. 
E. Launert : various. 



J. Leonard : Euphorbiaceae. 

Maria Candida Liberate : Leguminosae. 

J. B. Marshall : Compositae. 

E. Milne-Redhead : various. 

D. Napper : Acanthaceae. 

R. M. Polhill : Leguminosae. 

J. Raynal : Cyperaceae. 

N. K. B. Robson : Hypericaceae. 

R. Ross : for much help throughout. 

W. W. Sanford : Orchidaceae. 

P. Taylor : Amaranthaceae. 



NEW NAMES PUBLISHED IN THIS WORK 

Pauridiantha camposii (G. Tayl.) Exell, comb. nov. (Rubiaceae). 
Psychotria hierniana Exell, nom. nov. (Rubiaceae). 



DICOTYLEDONES 



I. ANNONACEAE 



Annona arenaria Thonn. 1 

♦Anngna cherimola L. 25 

♦Annona glabra L. i, 10 

♦Anngna muricata L. I, 2, 16, 25, 26 

♦Annona reticulata L. i, 16, 25 

♦Annona squamosa L. i, 25 

Artabotrys macrophyllus Hook. f. 2, 10 
Artabgtrys sp. 

FP : Mildbraed 6412 

♦Asimina triloba (L.) Dunal i 

♦Cananga gdgrata (Lam.) Hook. f. & Thomson 2, 11, 25 

Cleistgphglis patens (Benth.) Engl. & Diels 2 

FP : Bokoko, Burchardt. Not recorded from FP in 10. 

Enantia chlorantha Oliver 10 

Mgngdora brevipes Benth. i, 2, 10 

Mgngdora myristica (Gaertn.) Dunal i, 2, 7, 25, 26 

Not recorded from FP in 10. 

Mgngdora tenuifolia Benth. 2, 10 

Mgngdora zenkeri Engl. 2 

Not in 10. 



FP 



FP 
FP 



FP 
FP 



ST 
ST 
ST 
ST 
ST 
ST 



? 
ST 



FP 

FPP 

FPPST 

FP 
FP 



lO 

2 






FP 
FP 


2, IC 






FP 


I. 2 






ST 
FP 


27 






ST 


lO 






FP 


2 








lO 






FP 


2 








lO 






FP 


2 








I. 2, 


25. 


26 


PST 


I, 2, 


lO 




FPPST 


I, 2, 


lO 




FP ST 



OF THE GULF OF GUINEA 337 

OXYMITRA GRACILIPES Benth. 

OxYMiTRA GRACILIS (Hook. f.) Spraguc & Hutch. 
Unona millenii Engl. & Diels 
Not recorded from FP in 10. 
OxYMiTRA HiRSUTA (Benth.) Sprague & Hutch. 

POLYALTHIA OLIVERI Engl. 
POPOWIA Sp. 

FP : Bokoko, Mildbraed 686g 

UVARIA Sp. 

UvARiODENDRON CONNIVENS (Benth.) R. E. Fries 

Uvaria connivens Benth. 
UvARiODENDRON FUSCUM (Benth.) R. E. Fries 

Uvaria fusca Benth. 
UVARIODENDRON LETESTUi (Pellegrin) R. E. Fries 

Uvaria mannii Hutch. & Dalz. 
Xylopia aethiopica (Dunal) A. Rich. 
Xylopia africana (Benth.) Oliver 

FP : Moka, Mildbraed 7023. Not recorded from FP 
in 10. 
Xylopia quintasii Engl. & Diels 
X. striata Engl. 
Not recorded from FP in 10. 

2. monimiaceae 

Glossocalyx longicuspis Benth. 

Xymalos monospora (Harv.) Baillon ex Warb. 

3. lauraceae 

♦Cinnamomum burmanni (C. & T. Nees) Blume 
♦Cinnamomum camphora (L.) T. Nees & Eberm. 
♦Cinnamomum zeylanicum Garc. ex Blume 
♦Persea AMERICANA Miller 
P. gratissima Gaertn. f. 

4. hernandiaceae 
Hernandia beninensis Welw. ex Henriq. 

5. MYRISTICACEAE 

Coelocaryon preussii Warb. FP 

FP : Bokoko, Mildbraed 6856. Not recorded from FP 

in 10. 
♦Myristica fragrans Houtt. 
Pycnanthus angolensis (Welw.) Warb. 
P. kombo (Baillon) Warb. 

Not recorded from FP in 10. 
Pycnanthus microcephalus (Benth.) Warb. 
Staudtia kamerunensis Warb. 

FP : Bokoko, Mildbraed 6883, on which the record in 

2 is probably based. Not recorded from FP in 10. 
Staudtia pterocarpa (Warb.) Warb. i, 25 ST 

Staudtia stipitata Warb. FP 

5. gabonensis Warb. 2 

Not recorded from FP in 10. 



2, 10 




FP 


2, 10 




FP 


I 




ST 


I. 25 




ST 


I. 2, 13 




PST 


I, 13. 16. 


25 


FP P ST A 


2 






I, 2, 7, 10 


FP ST 



I 




ST 


I, 10, II 

2 


FP 


ST 


2, 10 


FP 




2 


FP 





338 ANGIOSPERMS OF THE ISLANDS 

6. RANUNCULACEAE 

Clematis hirsuta Guillemin & Perrottet 8, ii, 15 ST 

Clematis simensis Fresen. 10 FP 

C. altissima Hutch. 2 

Ranunculus extensus (Hook, f.) Schube ex Engl. 2, 10 FP 

' Rubus extensus Schube ex Engl.', Index Kewensis, 

Suppl. I : 370 (1906), cited from FP but entirely 

mythical ; Ranunculus extensus was undoubtedly 

intended. 
Thalictrum rhynchocarpum Dillon & Rich. 10 FP 

T. mannii Hutch. 2, 7 

7. nymphaeaceae 

Nymphaea lotus L. 10 FP 

N. lotus var. sinoeensis Stapf 2 

8. menispermaceae 

Chasmanthera dependens Hochst 27 ST 

CissAMPELOs owARiENSis Beauv. ex DC 10 FP 

DiOSCOREOPHYLLUM VOLKENSII Var. FERNANDENSE 

(Hutch. & Dalziel) Troupin 12 FP 

D. fernandense Hutch. & Dalziel 2 
D. tenerum var. fernandense (Hutch. & Dalziel) 

Troupin 10 

Jateorhiza macrantha (Hook, f.) Exell & Mendon9a 2, 10 FP 

Penianthus longifolius Miers 2, 10 FP 

Stephania abyssinica (Dillon & Rich.) Walp. var. 

abyssinica 

Stephania cyanantha Welw. ex Hiern 12 FP 

Stephania dinklagei (Engl.) Diels 15 P 

Stephania laetificata (Miers) Oliver 2 FP 

Not recorded from FP in 10. The holotype, Mann 

236, is from FP. 
Syrrheonema fasciculatum Miers 2, 10 FP 

Triclisia coriacea Oliver 2, 10 FP 

Triclisia macrophylla OHver 2, 10 FP 

Recorded as endemic to FP in 10, but the distribytion 

has since been found to extend to Sierra Leone and 

Cameroons Mt. 

9. aristolochiaceae 

Aristolochia littoralis Parodi ST 

A. elegans Mast. 
Pararistolochia goldieana (Hook, f .) Hutch. & Dalziel 

10. PIPERACEAE 

Peperomia fernandopoiana CDC. 
Peperomia hygrophila Engl. 

p. annobonensis Mildbr. 
Peperomia kamerunana CDC 
'Peperomia lucida' 

Probably P. pellucida was intended. 



2, 10 FP 



II 

2, 10 


FP 


2, 7, 10, 15, 19 
I, 10, 16 


FPP 

ST A 


I, 2 
10, 19 

7 


FP 
FP 



OF THE GULF OF GUINEA 



339 



Peperomia mannii Hook. f. ex CDC. 
Peperomia molleri CDC. 
Peperomia pellucida (L.) Kunth 

Peperomia reflexa (L.f.) A. Dietr. 

Peperomia thomeana CDC 

Peperomia vaccinifolia CDC 

Peperomia sp. 

Piper capense L.f. 

Piper guineense Schumacher 

P. clusii CDC 
Piper umbellatum L. 

P. suhpeltatum CDC 



2, ID, ig 


FP 


I, ID, 19 


FPPST 


I, 2, 7, ID, II, 


FPPST 


19, 25, 26 




I. 15 


ST 


I 


ST 


ID, 19 


FP 


I 


A 


I, 2, 10, 11, 19 


FPPST 


I, 2, 7, ID, 19, 


FPPST 


25 




2 




I, 10, II, 19, 


FP P ST A 


25, 26 




2 





♦Argemone mexicana L. 



II. papaveraceae 



I, 25, 26 



ST 



12. fumariaceae 



*Fumaria muralis Bonder ex Koch 



ST 



13. capparaceae 



Buchholzia coriacea Engl. 

B. macrophylla sensu Espirito Santo 
Capparis tomentosa Lam. 

Cleome rutidosperma DC 

C. ciliata Schumacher 

Gynandropsis gynandra (L.) Briq. 
Pedicellaria pentaphylla Schrank 
Not recorded from FP in 10. 

RlTCHIEA ERECTA Hook. f. 
RlTCHIEA POLYPETALA Hook. f. 

R. fragrans sensu Mildbr. ? 
R. cf. glossopetala Gilg 
FP : Bokoko, Burchardt. 

R. obanensis Hutch. & Dalziel 



27 






ST 


25 








I 






ST 


16 






FP P ST A 


I, 2, 9, 


10, 


II, 




25 








I 






FP ST 


2 








2, 10 






FP 


10 






FP 



14. CRUCIFERAE 

♦Brassica juncea (L.) Czerniak 
*Brassica oleracea L. 
♦Capsella bursa-pastoris (L.) Medik. 
Cardamine africana L. 

C africana var. pubescens Hook. f. 
Cardamine hirsuta L. 
♦CoRONOPus DiDYMUs (L.) Smith 
*DiPLOTAxis tenuisiliqua Dehle 

♦LOBULARIA MARITIMA (L.) DcSVaUX 

*Raphanus sativus L. 

RoRippA HUMiFUSA (Guillcmin & Perrottet) Hiern 
Nasttirtium humifusum Guillemin & Perrottet 
Not recorded from FP in 10. 



1, 15 




ST 


2 


FP 




I 




ST 


I, 10, 13 


FP 


ST 


2 






2, 7, 10 




ST 


I, II 




ST 


I 




ST 


15 




ST 


I. 13 




ST 



FP 



340 ANGIOSPERMS OF THE ISLANDS 

RoRiPPA iNDiCA (L.) Hiern i6 ST 

Nasturtium sinapis (Burm. f.) O. E. Schulz ii, 14 

Rorippa sinapis (Burm. f.) Ohwi & Hara 15 

♦RORIPPA NASTURTiuM-AQUATicuM (L.) Hayek 15, 25, 26 FP ST 

Nasturtium officinale R.Br. i, 2 

♦Sinapis alba L. 2 FP 

15. violaceae 

Hybanthus enneaspermus (L.) F. v. Muell. 10 FP 

RiNOREA ANGUSTiFOLiA (Thouars) BaiUon ST A 

R. molleri Brandt i, 9, n 

R. ardisiiflora sensu Guinea 2 

See Tennant in Kew Bull. 16 : 412 (1962). 

RiNOREA CHEVALIERI Exell I ST 

RiNOREA DENTATA (Beauv.) Kuntze 2, 7, 10 FP 

R. bipindensis Engl. 2 

RiNOREA INSULARIS Engl. I P 

RiNOREA LONGISEPALA Engl. 2 FP 

Not recorded from FP in lo. 

RiNOREA THOMENSIS Exell I ST 

Viola abyssinica Steudel ex Oliver 2, 10 FP 

V. emirnensis (Engl.) Guinea 2 

16. polygalaceae 

Carpolobia glabrescens Hutch. & Dalziel 2, 10 FP 

17. CRASSULACEAE 

♦Bryophyllum pinnatum (Lam.) Oken i, 11, 25, 26 FP P ST A 

B. germinans sensu Guinea 2 

Crassula alsinoides (Hook, f.) Engl. 2, 7 FP 

Not recorded from FP in 10. The type is from FP. 
Kalanchoe crenata (Andr.) Haw. i, 10, 11, 16, 27 ST A 

Kalanchoe crenata var. verea (Jacq.) Cufod. in Bull. 

Jard. Bot. Brux. 27 : 714 (1957) • ST 

ST : Menyharth 24 (WU). 
Kalanchoe laciniata (L.) DC. FP 

FP : Moka, Exell yg6 (BM.). Not recorded from FP 

in 10. 

18. hydrangeaceae 

♦Hydrangea sp. i ST ? 

19. podostemaceae 

Tristicha trifaria (Bory) Sprengel 10 FP 

T. hypnoides (St.-Hil.) Sprengel 2 

20. CARYOPHYLLACEAE 

Cerastium indicum Wight & Arn. 10 FP 

Drymaria cordata (L.) Willd. i. 2, 8, 10, 11, FP P ST 

13. 25 

Sagina abyssinica Hochst. ex A. Rich. 2, 10 FP 

Stellaria mannii Hook. f. i, 10, 11, 13 FP ST 

♦Stellaria media (L.) Vill. i FP ST 
FP : Guinea 20$j. 



I 



I 

I 



OF THE GULF OF GUINEA 

21. AIZOACEAE 



lESUVIUM PORTULACASTRUM (L.) L. 



PORTULACA OLERACEA L. 



ST 



22. PORTULACACEAE 



PORTULACA QUADRIFIDA L. 

*Talinum triangulare (Jacq.) Willd. 

23. POLYGON ACEAE 

•Antigonon leptopus Hook. & Am. 

A. cordatum Martens & Galeotti 
Polygonum nepalense Meissner 

P. alatum (Buch.-Ham. ex D. Don) Sprengel 
Polygonum nyikense Baker 
Polygonum salicifolium Brouss. ex Willd. 
Polygonum senegalense Meissner 
RuMEX abyssinicus Jacq. 
Not recorded from FP in 10. 



24. PHYTOLACCACEAE 

Phytolacca dodecandra var. apiculata Baker & 

Wright I, II, 13 



I, 2, 


10, 


13. 


15. 


FP 


ST 


16, 


25 










13 










ST 


I, 2, 


II, 


16, 


25 


FP 
FP 


ST 
ST 


2, 13 












10 








FP 




2 












10 








FP 




I 












I 










ST 


I, 2 








FP 


ST 



ST 



ChENOPODIUM ALBUM L. 
♦Chenopodium AMBROSIOIDES L. 



25. chenopodiaceae 



26. amaranthaceae 



ACHYRANTHES ASPERA L. 

Not recorded from FP in 10. 

ACHYRANTHES BIDENTATA Blume 

Aerva lanata (L.) Juss. ex J. A. Schultes 

FP : Guinea 112 j. Not recorded from FP in 10. 
Alternanthera maritima (Mart.) St.-Hil. 
Telanthera maritima (Mart.) Moq. 

FP : Guinea 875. Not recorded from FP in 10. 
Alternanthera pungens Kunth 

A. repens (L.) Link, non A. repens J. F. Gmelin 
Alternanthera sessilis (L.) R. Br. ex DC, Cat. Hort. 

Monspel. : 77 (1813). 
Amaranthus graecizans L. 
A. blitum sensu auctt. 

The record in 13 is probably A. graecizans. 
Amaranthus hybridus subsp. incurvatus (Timeroy ex 

Gren. & Godron) Brenan 
A . paniculatus L. 

Not in 10. 
Amaranthus spinosus L. 
Amaranthus viridis L. 

Not recorded from FP in 10. 
*Celosia argentea L. 

Probably introduced. 



I 




ST 


I, II, 13, 25 


26 


PST 


I, 2, 16, 25 




FP ST A 


2, 10 




FP 


2 {' Aurea') 




FP 


16 




FP P ST A 


I, 14 




PST 


I, II, 13 






I, 8, 10, II, 


16 


FP P ST A 
ST 


13 







I, 2, 10, 13, 16 
I, 2, 16 

I, 2 



FP 



FP P ST A 
FP ST A 

FP ST 



341 



342 



ANGIOSPERMS OF THE ISLANDS 



Celosia bonnivairii Schinz 

Celosia laxa Schumacher 

Celosia leptostachya Benth. 

Celosia trigyna L. 

Cyathula achyranthoides (Kunth) Moq. 

Cyathula cylindrica var. mannii (Baker) Suesseng. 

C. mannii Baker 
Cyathula fernandopoensis Suesseng. & Friedrich 
Cyathula pedicellata C. B. Clarke 
Cyathula prostrata (L.) Blume 
Deeringia amaranthoides (Lam.) Merr. 

♦GOMPHRENA GLOBOSA L. 

♦Iresine herbstii Hook. 

NOTHOSAERVA BRACHIATA (L.) Wight 

FP : EscarrS 361, jyj, 373. Not recorded from FP 
in 10. 
Philoxerus vermicularis (L.) Beauv. 

FP : Escarrd 360. Not recorded from FP in 10. 

PUPALIA LAPPACEA (L.) JuSS. 

Not recorded from FP in 10. 
Sericostachys scandens Gilg & Lopr. 

27. basellaceae 
*Basella alba L. 

Probably introduced. 

28. GERANIACEAE 

Geranium arabicum Forssk. 

G. simense Hochst. ex A. Rich. 
See Laundon in Bol. Soc. Brot., ser. 2, 35 : 59 (1961). 



OXALIS corniculata L. 



♦OXALIS CORYMBOSA DC. 



♦Averrhoa carambola L. 



29. oxalidaceae 



30. averrhoaceae 



31. balsaminaceae 
♦Impatiens balsamina L. 
Impatiens buccinalis Hook. f. 
Impatiens filicornu Hook. f. 

(Omitted from index to 10 ; text p. 162.) 
Impatiens hians Hook. f. 
Impatiens kamerunensis Warb. 
Impatiens macroptera Hook. f. 
Impatiens mannii Hook. f. 

/. affinis Warb. 
Impatiens manteroana Exell 
Impatiens niamniamensis Gilg 

I. dichroa Hook. f. ex Gilg 
Impatiens sakerana Hook. f. 
Impatiens tamsiana Exell 
Impatiens thomensis Exell 



10 






FP 




2, 10 






FP 




I, 2, 10 


, II 


. 13 


FP 


ST 


I, 10 






FP 


A 


2, 10 






FP 




10 






FP 




2 










10 






FP 




10 






FP 




I, 10, I 


3. I 


6 


FP P ST A 


I 






P 




I, 2 






FP 


ST 


I 






FP 


ST 


I, 19 






FP 


ST 


I, 2, 9, 


II, 


25 


FP 


ST A 


10 






FP 





I, 16 



2, 7, 10 



ST A 



FP 



I, 2, 10, 


II, 12, 


FP P ST A 


13 
I, II, 


13, 


25 




PST 


2, 13, 


25 




FP 


ST 


I, 13 








ST 


I, II 








ST 


2, 10 






FP 




2, 10 






FP 




10 






FP 




2, 10 






FP 




2, 10 






FP 




2 










I. 14 
10 






FP 


P 


2 










10 






FP 




I 








P 


I, 14 








ST 



J 



14 




ST 


I 






M. 15 




PST 


I. 8, 13 






14 




ST 


I 








FP 


ST 


I 






14 







OF THE GULF OF GUINEA 343 

32. lythraceae 
I^Lawsonia inermis L. ii P 

33. punicaceae 

*punica granatum l. 2, i3, i6 fp st a 

34. onagraceae 

LuDWiGiA ERECTA (L.) Hara 

Jussiaea erecta L. 
I.UDWIGIA HYSSOPIFOLIA (G. Don) Exell 

Jussiaea linifolia Vahl 
LuDwiGiA LEPTOCARPA (Nutt.) Hara 

Jussiaea leptocarpa Nutt. 
LuDwiGiA ocTovALvis (Jacq.) Raven subsp. octovalvis 
Jussiaea suffruticosa L. 
Ludwigia pubescens (L.) Hara 
FP : San Carlos, Mildbraed 6y8i. 

Neither Ludwigia nor Jussiaea is recorded from FP 
in 10. 

35. HALORAGACEAE 

♦Laurembergia cf. TETRANDRA (Schott) Kanitz 10 FP 

36. THYMELAEACEAE 

DiCRANOLEPis DisTiCHA Planchon 

D. mannii Baillon 
DiCRANOLEPis grandiflora Engl. 
DiCRANOLEPis STENURA Gilg ex Engl. 
Not in 10, but probably = D. vestita. 

DiCRANOLEPis THOMENSIS Engl. & Gilg 
DiCRANOLEPis VESTITA Engl. 

Peddiea fischeri Engl. 

FP : Melville S' Wrigley 632 (K.). Not recorded from 

FP in 10. 
Peddiea parviflora Hook. f. 
Peddiea thomensis Exell 

37. nyctaginaceae 

BoERHAviA coccinea Miller 

B. diffusa sensu Exell pro parte et sensu Lains & 
Silva. 
BoERHAviA coccinea var. viscosA (Lag. & Rodriguez) 

Moscoso FP 

B. viscosa Lag. & Rodriguez 9. 

Not recorded from FP in 10. 

BoERHAVIA diffusa L. 
♦BOERHAVIA PANICULATA Rich. 

♦Bougainvillea glabra Choisy 

♦BOUGAINVILLEA SPECTABILIS WiUd. 
*MlRABILIS JALAPA L. 

38. PROTEACEAE 

♦Grevillea robusta Cunn. 13 ST 



10 




FP 






2 










2, 10 




FP 






2 




FP 




E? 


I 






ST 


E 


2, 10 




FP 
FP 






2, 10 




FP 




E 


I. 13 






ST 


E 


16 






PST A 




I, 13, 


26 









I, 16, 25, 26 


PST A 


I 


ST 


2 


FP 


I, 2, 13 


FPPST 


I, 2, II, 26 


FP ST A 



344 ANGIOSPERMS OF THE ISLANDS 

39. DILLENIACEAE 
♦DiLLENIA INDICA L. I, 27 ST 

Tetracera alnifolia Willd. i, 2, 7, 10, 16 FP P A 

Tetracera eriantha (Oliver) Hutch. \ 2, 10 FP 

40. pittosporaceae 

PiTTOSPORUM MANNII Hook. f. Subsp. MANNII 2, lO FP 

41. BIXACEAE 
*BlXA ORELLANA L. 



42. FLACOURTIACEAE (incl. SAMYDACEAE) 

Caloncoba glauca (Beauv.) Gilg 
♦Caloncoba welwitschii (Oliver) Gilg 
Camptostylus MANNII (Oliver) Gilg 
Caloncoba mannii (Oliver) Gilg 

CaSEARIA BARTER! Mast. 

Casearia MANNII Mast. 

Flacourtia flavescens Willd. 

? HoMALiUM AFRiCANUM (Hook. f.) Benth. 

Record probably based on a misidentification of H. 

letestui. 
HoMALiuM HENRiQUESii Gilg ex Engl. 

H. africanum sensu Exell 
HoMALiUM LETESTUI Pellegrin 

H. dolichophyllum Gilg ex Engl. 
HoMALiUM SARCOPETALUM Pierre 
LiNDACKERiA DENTATA (Oliver) Gilg 

FP : Bokoko, Burchardt. Not recorded from FP in lo. 
Oncoba spinosa Forssk. i, 2, 25, 26 FP ST 

Not recorded from FP in 10. 
Ophiobotrys zenkeri Gilg 

43. PASSIFLORACEAE 

Adenia cissampeloides (Planchon ex Benth.) Harms 

Adenia cynanchifolia (Benth.) Harms 

Adenia lobata (Jacq.) Engl. 

Barteria fistulosa Mast. 

Barteria nigritana Hook. f. 

♦Passiflora edulis Sims 

♦Passiflora foetida L. 

♦Passiflora quadrangularis L. 

44. cucurbitaceae 

General reference : C. Jeffrey, ' Key to the Cucurbitaceae of West Tropical Africa, with a guide 
to the localities of rare and little-known species' in Journ. W. Afr. Sci. Assn. 9 : 79-97 (1965). 
Cayaponia AFRICAN a (Hook. f.) Exell I, 9, 10, II ST A 

CiTRULLUS LANATUS (Thunb.) Matsum. & Nakai FP P ? 

C. vulgaris Schrader 2 

Colocynthis citrullus (L.) Kuntze I 

Not recorded from FP in 10. 



I, 9, 


10, II, 13, 


FPPST 


25. 


26 




)ACEA 
10 


E) 


FP 


2 




FP 


10 




FP 


2 






I, II, 

I 


. 25 


ST 
P 


II 




ST 


2 




FP? 


II 




ST 


I 






10 




FP 


2. 7 
10 




FP 
FP 



II, 25, 27 




ST 


2, 10, 26 




FP P ? ST ? 


2, 10 




FP 


I, 2, 10, II 




FP P ST A 


2, 10 




FP 


10 




FP 


I. II, 25 




ST 


I, 8, 10, 25, 


26 


FPPST 


I 




PST 



KocciNiA BARTERi (Hook. f.) Keay 

■ Physedra barteri (Hook, f.) Cogn. 
fcucuMEROPSis EDULis (Hook. f.) Cogn. 

■ Not recorded from FP in lo. 
KucuMis METULiFER E. Meyer ex Naudin 

^CUCUMIS SATIVUS L. 

♦CucuRBiTA MAXIMA Duchesne 
DiPLOCYCLos PALMATUS (L.) C. Jeffrey 

Bryonopsis laciniosa Naudin pro parte 
Lagenaria breviflora (Benth.) Roberty 

Adenopus brevifiorus Benth. 
Not recorded from FP in lo. 
Lagenaria siceraria (Molina) Standley 

L. vulgaris Ser. 

♦LUFFA ACUTANGULA Roxb. 

LuFFA AEGYPTiACA Miller 

L. cylindrica M. Roemer 
MoMORDicA CABRAE (Cogn.) C. Jeffrey 

Dimorphochlamys mannii Hook, f . 

MOMORDICA CHARANTIA L. 

MOMORDICA CHARANTIA var. ABBREVIATA Ser. 

MoMORDiCA cissoiDES Planchon ex Benth. 
MoMORDiCA FOETiDA Schumacher 
M. mannii Hook. f. 

MOMORDICA MULTIFLORA Hook. f. 

Oreosyce africana Hook. f. 
Peponium vogelii (Hook, f .) Engl. 

P. bracteatum (Cogn.) Cogn. 
Raphidiocystis mannii Hook. f. 
RuTHALiciA longipes (Hook. f.) C. Jeffrey 

Physedra longipes Hook. f. 
♦Sechium edule Swartz 
Telfairia occidentalis Hook. f. 
Zehneria capillacea (Schumacher) C. Jeffrey 

Melothria capillacea (Schumacher) Cogn. 
Zehneria gilletii (De Wild.) C. Jeffrey 

Melothria capillacea sensu Exell 
Zehneria minutiflora (Cogn.) C. Jeffrey 

FP : Wrigley 516 (K). Not recorded from FP in 10. 
Zehneria scabra (L. f.) Sonder 

Melothria minutiflora sensu Exell 

M. fernandensis Hutch. & Dalziel 



OF THE GULF OF GUINEA 345 



45. BEGONIACEAE 



Begonia ampla Hook. f. 



Begonia annobonensis A. DC. 
Begonia baccata Hook. f. 
B. cr uteris Exell 

I agree that B. crateris is no more than a form of B. 

baccata. 
Begonia eminii Warb. 
Begonia fusialata Warb. 



10 


FP 




2 






2 


FP 




I 




ST? 


I 




ST 


2 


FP 




16 


FP 


ST 


I, 10 






16 


FP P ST A 


I. 2, 25 






16, 25 


FP 


ST A 


2, 13 

2 


FP 




1, 10, 16 


FP 


ST A 


2 


FP 




2, ID 






I, 2, 9, 10, 13, 


FP P ST A 


25, 26 
I. 9. II 


FP P ST A 


2, 10 


FP 




2, 10 


FP 




2 






2, 10 


FP 




2, 10 


FP 




2, 10, 16 


FP 


ST 


I 






2, 10 


FP 
FP 




10 






I, 2 


FP 


ST 


2, 10 


FP 
FP 




2, 10 






16 




ST 


I 


FP 




16 


FP 


ST 


I 






2, 10 






I, 2, 7, 10, 23, 


FP P ST A 


25 
I, 2, 10, II, 23 




PST A 


I, II, 23 




ST 


I. 13. 25 






10 


FP 




23 




P 



346 



ANGIOSPERMS OF THE ISLANDS 



Begonia furfuracea Hook. f. 
Begonia gilgii Engl. 
Begonia gladiifolia Engl. 
Begonia loranthoides Hook. f. 

B. henriquesii CDC. 
Begonia macambrarensis Exell 
Begonia mannii Hook. 

B. epiphytica Hook. f. 

B. excelsa Hook. f. 
Begonia molleri (CDC.) Warb. 
Begonia oxyloba Welw. ex Hook. f. 
Begonia poculifera Hook. f. 
Begonia prismatocarpa Hook. 
Begonia sessilifolia Hook. f. 
Begonia thomeana CDC 
Begonia sp. i 
Begonia sp. 2 
Begonia sp. 3 
Begonia sp. 4 
♦Begonia sp. 
Begonia sp. cf. macropoda Gilg 



♦Carica papaya L. 



46. caricaceae 



47. cactaceae 

♦Opuntia dillenii (Ker-Gawl.) Haw. 

O. horrida Salm-Dyck 
Rhipsalis baccifera (J. S. Miller) Steam 

R. cassutha Gaertn. 

48. OCHNACEAE 

Ochna membranacea Oliver 

O. quintasii (Tieghem) Exell 
O. cf. gilgiana Engl. 
Ouratea affinis (Hook, f.) Engl. 
? Ouratea brachybotrys Gilg 

FP : Bokoko, Mildbraed 6866. A doubtful record. 
Not in 10. 
Ouratea brunneopurpurea Gilg 

Not in 10. 
Ouratea calophylloides Hutch. & Dalziel 
Ouratea elongata (Oliver) Engl. 
Ouratea flava (Schumacher) Hutch. & Dalziel ex Stapf 

O. reticulata sensu Guinea 
Ouratea mannii (Oliver) Engl. 
Ouratea micrantha (Hook, f.) Hutch. & Dalziel 
Ouratea molleri (Tieghem) Exell 

Campy lospermum vogelii var. molleri (Tieghem) 
Farron in Bull. Jard. Bot. Brux. 35 : 404 (1965). 
Ouratea nutans (Hiern) Exell 
Ouratea quintasii (Tieghem) Exell 

Rhahdophyllum arnoldianum var. quintasii (Tieghem) 
Farron, torn. cit. : 391 (1965). 



2, 10 


FP 




E 


2, 10 


FP 






2, 10 


FP 






I, 23 




PST 


E 


I 








I, 13. 23 




ST 


E i 


2, 10 


FP 






2 








2 








I, 13. 23 




ST 


E 


2, 10 


FP 






2, 10 


FP 






2, 10 


FP 




4 


2, 10 


FP 






I 




ST 


E 


I 




P 




I 




ST 




X 




ST 




I 




ST 




II, 23 




ST 




10 


FP 







1, 16, 25 



2 

11 

I, 2, ID 



16 

1. 13 

I 

2, 10 
2 



Ii 2, 14 



2, 10 




2, 10 




10 




2 

2, 10 




2, 10 




I, II. 


13 


I. II 




I. 13 





PST A 

FP 

FP P ST A 

ST A 



FP 
FP 



FPPST 

FP 

FP 
FP 
FP 
FP 

FP 

PST 



E 
E? 



ST 



E 

E? 



OF THE GULF OF GUINEA 



347 



49. MYRTACEAE 

♦Eucalyptus robustus Smith 

♦Eucalyptus sp. 

♦Eugenia brasiliensis Lam. 

♦Eugenia dombeyi (Sprengel) Skeels 

Eugenia elliotii Engl. & Brehmer 

Eugenia fernandopoana Engl. & Brehmer 

Eugenia obanensis Baker f. 

♦Eugenia uniflora L. 

♦Myrciaria cauliflora O. Berg 

♦PSIDIUM guajava L. 

♦Syzygium aromaticum (L.) Merr. & Perry 

Syzygium guineense (Willd.) DC. 

Syzygium guineense var. littorale Keay 

♦Syzygium jambos (L.) Alston 

Eugenia jambos L. 
♦Syzygium malaccense (L.) Merr. & Perry 

Eugenia malaccensis L. 
Syzygium staudtii (Engl.) Mildbr. 
ST : Monod iig2g. 



2 






FP 




13 








ST 


13 








ST 


27 








ST 


II 








ST 


2, 10 






FP 




2, 10 






FP 




I, 2, 25 






FPPST 


13 








ST 


I. 2, 13, 


25. 


26 


FP P ST A 


27 








ST 


2, 15. 25 


, 26 


. 27 


FP 


ST 


10 






FP 










FP P ST A 


I, 2 






FP 




2 










10 






FP 


ST 



Napoleonaea mannii Miers 
N. cuspidata sensu Guinea 



50. barringtoniaceae 



FP 



51. melastomataceae 

Amphiblemma cymosum var. bolivari Brenan & Guinea, 
nom. provis. 

A. Fernandes (in litt.) suggests that this may be A. 
mildbr aedii. 
Amphiblemma mildbraedii Gilg ex Engl. 
Calvoa confertifolia Exell 
Calvoa crassinoda Hook. f. & Triana 
Calvoa grandifolia Cogn. 
Calvoa henriquesii Cogn. 
Calvoa hirsuta Hook. f. 
Calvoa integrifolia Cogn. 
Calvoa molleri Gilg 
Calvoa sinuata Hook. f. ex Triana 
Calvoa aff. trochainii Jacques-F61ix 

C. ignatii-bolivari Guinea, nom. provis. 

C. polychoeta Guinea, nom. invalid. 

Referred to in the legend of the plate on p. 179 

but not in the text, and presumably the same as C. 

ignatii-bolivari. 
Calvoa uropetala Mildbr. 

DiCELLANDRA BARTERI Hook. f . 

Dicellandra scandens Gilg ex Engl. 
DiNOPHORA spenneroides Benth. 
DissoTis BARTERI Hook. f. ex Triana 
DissoTis DECUMBENS (Beauv.) Triana 
DissoTis rotundifolia (Smith) Triana 



FP 



2, 10 




FP 




1. 24 




ST 


E 


1. 13. 24 




ST 


E 


I, 8, 13, 24 




ST 


E 


I, II, 13, 15, 


24 


PST 


E 


2, 5. 10 




FP 




I. 24 




ST 


E 


I, II, 15, 24 




ST 


E 


I, 24 




P 


E 


5. 10 




FP 




5 








5 









I, 
2, 
2, 


2 

5. 7. 
10 


10 




FP 
FP 


A. 


E 
E 


2, 


10 






FP 






I, 


24 






P 




E 


2, 


5. 10 






FP 






I, 


5. 7. 


10, 


24 


FPPST 







348 



ANGIOSPERMS OF THE ISLANDS 



GUYONIA CILIATA Hook. f. 

Afzeliella holivari Brenan & Guinea, nom. provis. 

Afzeliella sp. aff. ciliata 
Medinilla mannii Hook. f. 

Memecylon fernandianum Gilg ex Engl. \ 

Memecylon membranifolium Hook. f. 
Memecylon memecyloides (Benth.) Exell 

M. vogelii Naudin 
Memecylon aff. strychnoides Baker 
Memecylon tessmannii Gilg ex Engl. 

Not in lo. 
Myrianthemum mirabile Gilg 
Tristemma coronatum Benth. 
Tristemma hirtum Beauv. 

Tristemma incompletum R. Br. 

Tristemma littorale Benth. 
Tristemma mildbraedii Gilg ex Engl. 
Tristemma oreothamnus Mildbr. 
Tristemma radicans Gilg ex Engl. 
Tristemma thomensis Barros Ferreira 

52. COMBRETACEAE 

Combretum bracteatum (M. a. Lawson) Engl. & Diels 

CoMBRETUM coNFERTUM (Benth.) M. A. Lawson 

Combretum cuspidatum Planchon ex Benth. 

Combretum insulare Engl. & Diels 

Combretum paniculatum Vent. 

? Combretum platypterum (Welw.) Hutch. & Dalziel 

A doubtful record. 
Combretum racemosum Beauv. 

CONOCARPUS ERECTUS L. 

Not recorded from FP in 10. 
Laguncularia racemosa (L.) Gaertn. f. 

Not recorded from FP in 10. 
♦quisqualis indica l. 
♦Terminalia catappa L. 

53. rhizophoraceae 

Anisophyllea cabole Henriq. 
Anisophyllea setosa Mildbr., nom. nud. 

FP : Bokoko, Mildhraed 6836. 
Anisophyllea sp. 
Cassipourea annobonensis Mildbr. 
Cassipourea glomerata Alston 
Cassipourea gummiflua var. mannii (Hook. f. ex 

Oliver) J. Lewis 

C. mannii Hook. f. ex Oliver 
Rhizophora harrisonii Leechm. 

R. racemosa sensu Exell 
Rhizophora mangle L. 

This may be R. harrisonii. The genus is not recorded 

from FP in 10. 



10 








FP 








5 
















5 
2, 10 








FP 






E 


2, 10 








FP 






E 


2, 10 








FP 








I, 10, 


24. 


25 




FPP 








X 












A 




2 








FP 








2, 10 








FP 








24 
I, 2, 5 


10, 13, 


P 
FPP ST 




E 


14. 3 
1,5.8 


5. 
9. 


24 
10, 


II, 


FPP ST 




ij 


13. ] 

1. 2, 5 

2, 10, 


[4. 
7. 
24 


15. 
10, 


24 
24 


FPP 
FP 


ST 




1 


I. 2. 5 


16 








A 


E 


2, 10 








FP 






E 


24 










ST 




KJ 


10 








FP 






fl 


2, ID 








FP 








2, 10 








FP 








2, 10 








FP 








I, 2, 10, 


[I 




FPP 








I 








P 


? 






2, ID 








FP 








I. 7 








FP 


ST 







2, 7 



FP 



I 


FP ST 




I, 2, 16 


FP P ST A 




I. 13, 25 


ST 


E 


ID 


FP 


E? 


15 


P 




I, 2, 16 


A 


E 


I 


ST 
ST 


E 


I, 13 






II, 13. 15 


PST 




I 






7 


FP? 





OF THE GULF OF GUINEA 



349 



54. HYPERICACEAE 

Harungana MADAGASCARiENSis Lam. ex Poiret 

H. paniculata Pers. 
Hypericum peplidifolium A. Rich. 
Hypericum revolutum Vahl 

H. lanceolatum Lam. 

H. sp. 

55. guttiferae 

Allanblackia monticola Mildbr. ex Engl. 
Garcinia kola Heckel 

Not recorded from FP in 10. 
♦Garcinia mangostana L. 
Garcinia mannii Oliver 

Not recorded from FP in 10. 
Garcinia polyantha Oliver 
Garcinia sp. 

FP : Bokoko, Burchardt. 
Mammea africana Sabine 

Pentadesma butyraceum sensu Exell et al. 
Mammea klaineana Pierre 
? Pentadesma butyraceum Sabine 

See 'Antidesma butyraceum', p. 352. 
Symphonia globulifera L. f. 
S. gabonensis (Vesque) Pierre 

FP : Bokoko, Mildbr aed 6886, 6g65. Not recorded 

from FP in 10. 

56. scytopetalaceae 
Rhaptopetalum coriaceum Oliver 
Scytopetalum kamerunianum Engl. 

SCYTOPETALUM Sp. 

This may be 5. kamerunianum. 



I, 8, 


10, 


11, 


13. 


FPPST 


25 










2 










10 








FP 
FP 


2. 7. 


10 








10 










2, 10 








FP 


2 








FP 


13 








ST 


2 








FP 



I, 2, 10 


FPP? 




FP 


I, 10, 15, 27 


PST 


I, II, 13, 25 




2 






FP? 


10, 26, 27 


FP ST 


I 





2, 12 
25. 27 



FP 



ST 
ST 



♦Berria cordifolia (Willd.) Burret 
B. amomnilla Roxb. 

CORCHORUS AESTUANS L. 
CORCHORUS CAPSULARIS L. 

Not in 12. 

CoRCHORUS OLITORIUS L. 

Not recorded from FP in 12. 
DuBosciA MACROCARPA Bocquet 

Not recorded from FP in 12. 
Glyphaea brevis (Sprengel) Monachino 
Grewia carpinifolia Juss. 
Grewia floribunda Mast. 
Grewia malacocarpa Mast. 
Grewia oligoneura Sprague 
Triumfetta cordifolia a. Rich. 
Triumfetta eriophlebia Hook. f. 
Triumfetta rhomboidea Jacq. 

Not recorded from FP in 12. 



57. TILIACEAE 



FP 



ST A 



2 


FP 




I, 2 


FP 


ST 


2 


FP 




12 


FP 




I. 13 




ST 


I 




ST 


12 




ST 


2, 12 


FP 




2, 12 


FP 




2, 12 


FP 




I. 2 


FP 


ST A 



I. 2, 


12, 


13. 


15 


FP P 


2 

I 








P 


2, 12 








FP 


12 








FP 


2 










2 










2 








FP 



350 ANGIOSPERMS OF THE ISLANDS 

58. STERCULIACEAE 

Cola acuminata (Beauv.) Schott & Endl. i. 2, 12. 13, 15 FP P ST 

C. pseudoacuminata Engl. 
Cola digitata Mast. i P ST 

Cola ficifolia Mast. 
Cola nitida (Vent.) Schott & Endl. 

C. acuminata var. latifolia K. Schum. 
C. vera K. Schum. 
Cola obtusa Engl. & Krause 

Not in 12. 
? Cola sphaerocarpa A. Chev. i ST ? 

Said to be endemic but a very doubtful species. 
Cola verticillata (Thonn.) Stapf ex A. Chev. 2 FP 

Not recorded from FP in 12. 
Leptonychia adolfi-friederici Engl. & Krause 2, 12 FP E 

Leptonychia batangensis (C. H. Wright) Burret 12 FP 

Leptonychia densivenia Engl. & Krause 12 FP E 

L. urophylla sensu Guinea 2 

Leptonychia fernandopoana Engl. & Krause ex 

Mildbr., nom. nud. 12 FP E 

FP : Bokoko, Mildbr aed 6823, 6862. 
Leptonychia lanceolata Mast. 2, 12 FP E 

Leptonychia pallida K. Schum. 12 FP 

Melochia corchorifolia L. I ST 

♦Pterygota alata (Roxb.) R.Br. 2 FP 

? Sterculia bequaertii De Wild. 25 ST 

♦Sterculia foetida L. 2 

Sterculia oblonga Mast. 2, 

If the genus Eribroma Pierre be maintained, the name 

for this is E. oblonga (Mast.) Bodard 
Sterculia tragacantha Lindl. i. 

Not recorded from FP in 12. 
♦Theobroma cacao L. I, 

Waltheria indica L. 

FP : Wrigley & Melville 704 {K). Not recorded from 

FP in 12. 

59. bombacaceae 
adansonia digitata l. i, 9. 25, 26 fp p vst a 

A. digitata var. congolensis A. Chev. i» I3 

Not recorded from FP in 12. 
BoMBAx buonopozense Beauv. 2 FP 

Not recorded from FP in 12. 
*BoMBAX KiMUENZAE De Wild. & Dur. 2 FP 

Ceiba pentandra (L.) Gaertn. i, 2, 7, 13, 25 FP P ST A 

Not recorded from FP in 12. 

60. MALVACEAE 

♦Abelmoschus esculentus (L.) Moench 22 P ST 

Hibiscus esculentus L. i. 25 

Abelmoschus moschatus Medik. 2, 22 FP ST 

Hibiscus abelmoschus L. i 

Abutilon grandiflorum G. Don i, 22, 25, 26 ST 





FP 


12 


FP 


2 


FPPST 


2, II, 13, 25 


FP P ST A 




FP 



OF THE GULF OF GUINEA 



351 



►Abutilon grandifolium (Willd.) Sweet 
A. mollissimuni (Cav.) Sweet 
See R. Fernandas in Anu. Soc. Brot. 25 : 29 (1959). 
•■Abutilon venosum Lemaire 

►GOSSYPIUM BARBADENSE L. 

*var. BRASiLiENSE (Macfad.) J. B. Hutch. 
G. brasiliense Macfad. 
G. arboreum sensu Guinea 

^GOSSYPIUM HIRSUTUM L. 

■♦Hibiscus acetosella Welw. ex Hiern 
Hibiscus donianus D. Dietr. 
♦Hibiscus mutabilis L. 

Hibiscus physaloides Guillemin & Perrottet 
♦Hibiscus rosa-sinensis L. 
♦Hibiscus sabdariffa L. 
♦Hibiscus schizopetalus (Mast.) Hook. f. 

FP : Guinea 97. 
Hibiscus surattensis L. 

Not recorded from FP in 12. 
♦Hibiscus syriacus L. 
Hibiscus tiliaceus L. 

♦Malvastrum coromandelianum (L.) Garcke 
Pavonia urens Cav. var. urens 
Pavonia urens var. glabrescens (Ulbr.) Brenan 

P. schimperana sensu Guinea 
SiDA ACUTA Burm. f. 

Sida acuta var. carpinifolia K. Schum. 

SiDA ALBA L. 

FP : Moka, Tessmann 28 ly. Not recorded from FP 
in 12. 
SiDA cordifolia L. 

Not recorded from FP in 12. 

SiDA RHOMBIFOLIA L. 

SiDA RiGiDA (D. Don) D. Dietr. 

SiDA STIPULATA Cav. 

SiDA URENS L. 

SiDA VERONICIFOLIA Lam. 

Urena lobata L. 

WissADULA rostrata (Schumacher) Hook. f. 

W. amplissima var. rostrata (Schumacher) R. E. 

Fries 
W. heterosperma Hochst. ex Mast. 



ST 



I, 22 



I, 22 




ST 


I, 16, 22 




PST A 


12 




FP 


2 






2 






I, 22 




ST 


II, 15, 22, 


25 


PST 


I, 22 




ST 


15. 22 




PST 


I, 9, II, 22 


ST A 


I, 22 




PST 


2, 14, 22 




FP ST 
FP 


I, 2, 22 




FPPST 


2 




FP 


I, 2, 7, 12, 


22 


FP? ST 


15, 22 




PST 


12 




FP 


12 




FP 


2 






I, 8, 9, II, 


12, 


FP P ST A 


13. 22, 25 




2 




FP 


I, 22 




FP A 



I, 2, 9, II, 22 FP ST A 

I, 12, 13, 16, 22 FP P ST A 

I, 22 ST 

12 FP 

I, 22 ST 

I, 22 ST 

I, 9, II, 12, 13, FP P ST A 

22, 25 
15, 22, 25 FP P ST A 

I, 9, II, 12 
2 



61. malpighiaceae 

Acridocarpus longifolius (G. Don) Hook. f. i, 2, 12 

Flabellaria paniculata Cav. 2, 12 

Heteropteris leona (Cav.) Exell i, 12 



FP P ST 
FP 

PST 



62. HUMIRIACEAE 

Sacoglottis gabonensis (Baillon) Urban 



2, 12 



FP 



I, 2, l6 








I 






ST 


12 






FP 


2 






FP 


12 






FP 


2 








I, 12, 13. 


25. 


26 


FPPST 


2, 12 






FP 


I 






P 
ST 


I 








13 






ST 


I, II, 12 

7 






PST 
FP 



352 ANGIOSPERMS OF THE ISLANDS 

63. ERYTHROXYLACEAE 

♦Erythroxylum coca Lam. i ST 

Erythroxylum emarginatum Thonn. 9, 11, 27 ST A 

64. LINACEAE (hUGONIACEAE) 

HuGONiA platysepala Welw. ex Oliver 1,12 FP P 

HuGONiA SPiCATA Oliver var. spicata 2, 12 FP 

The type variety is endemic. 

65. euphorbiaceae 
ACALYPHA annobonae Pax & Hoffm. 

ACALYPHA CILIATA FOFSSk. 
ACALYPHA MANNIANA MiiU. Arg. 

AcALYPHA ornata Hochst. ex A. Rich. 

Not recorded from FP in 12. 
AcALYPHA racemosa Wall, ex Baillon 

A. paniculata Miq. 
Alchornea cordifolia (Schumacher) MiiU. Arg. 
Alchornea floribunda MiiU. Arg. 
Alchornea hirtella Benth. 
Alchornea laxiflora (Benth.) Pax & Hoffm. 

Lepidoturus laxiflorus Benth. 
♦Aleurites cordata (Thunb.) R. Br. ex Steudel 
Anthostema aubryanum Baillon 
'Antidesma butyraceum' 

Possibly Pentadesma butyraceum was intended. If so, 

it is a new record of that species from FP. 
Antidesma laciniatum MiiU. Arg. var. laciniatum 2, 12 FP 

Antidesma membranaceum MiiU. Arg. i ST A 

Antidesma oblonga (Hutch.) Keay FP 

FP : Wrigley &' Melville 633 (K). Not recorded from 

FP in 12. 
♦Breynia nivosa (W. G. Smith) Small 

Phyllanthus nivosus W. G. Smith 
Bridelia stenocarpa MiiU. Arg. 

B. micrantha sensu Exell 
Not recorded from FP in 12. 

Caperonia latifolia Pax 

Cavacoa quintasii (Pax & Hoffm.) J. L^nard 

Grosser a quintasii Pax & Hoffm. 
Cleistanthus libericus N. E. Brown 

C. polystachyus sensu Hutch. 
♦CoDiAEUM variegatum (L.) Blume 
♦Croton draconopsis MiiU. Arg. 
Croton lobatus L. 
Croton stellulifer Hutch. 

Crotonogyne MANNIANA MiiU. Arg. subsp. manniana 
Discoclaoxylon hexandrum (MiiU. Arg.) Pax & Hoffm. 

Claoxylon hexandrum MiiU. Arg. 
Discoclaoxylon occidentale (MiiU. Arg.) Pax & 

Hoffm. I, 2, 13, 16 ST E 

Discoclaoxylon pedicellare (MiiU. Arg.) Pax & 

Hoffm. 2, 16 FP E 

Claoxylon pedicellare MiiU. Arg. 12 



27 


ST 


13 




7. II. 25 


FPPST 


I, 2, 7 




I, 2, 12 


ST 


15 


ST 


1 




I, 16, 27 


ST 


II, 12, 25, 27 




I. 16 


ST 


I 


ST 


12 


FP 


I 


P 


2, 7, 12 


FP 


2, 16 


FP 


12 





OF THE GULF OF GUINEA 353 

DiscocLAOxYLON PUBESCENS (Pax & Hoffm.) Exell 
D. occidentale var. pubescens Pax & Hoffm. 
For a general discussion of Discoclaoxylon see 16. 
DiscoGLYPREMNA CALONEURA (Pax) Prain 
Drypetes fernandopoana Brenan 

D. glomerata (Miill. Arg.) Hutch. 
Drypetes glabra (Pax) Hutch. 
Drypetes henriquesii (Pax) Hutch. 
Drypetes laciniata (Pax) Hutch. 
Drypetes occidentalis (Miill. Arg.) Hutch. 
Drypetes principum (Miill. Arg.) Hutch. 
Elaeophorbia drupifera (Thonn.) Stapf 
Erythrococca anomala (Juss. ex Poiret) Prain 
Erythrococca columnaris (Miill. Arg.) Prain 
Erythrococca mannii (Hook, f.) Prain 

Athroandra mannii (Hook, f.) Pax & Hoffm. 
Erythrococca molleri (Pax) Prain 
Erythrococca pallidifolia (Pax & Hoffm.) Keay 

Athroandra pallidifolia Pax & Hoffm. 
Euphorbia cervicornu Baillon 
Euphorbia glaucophylla Poiret 

Not recorded from FP in 12. 
Euphorbia hirta L. 

Euphorbia hyssopifolia L. 

E. hypericifolia sensu Exell 
♦Euphorbia peplus L. 
Euphorbia prostrata Aiton 

♦Euphorbia pulcherrima Willd. ex Klotzsch 

Euphorbia schimperana G. H. A. Scheele FP 

E. ampla Hook, f . 2 

FP : Moka, Exell 845 (BM) . Not recorded from FP ini 2 . 
♦Euphorbia serpens Kunth 26 ST 

Euphorbia thymifolia L. i 

♦Euphorbia tirucalli L. 14 ST 

E. rhipsaloides Welw. ? i 

ExcoECARiA GuiNEENSis (Benth.) Miill. Arg. P 

Sapium guineense (Benth.) Kuntze i, 12 

See J. Leonard in Bull. Soc. Bot. Brux. 29 : 138 (1959). 
Grossera elongata Hutch. 
Grossera sp. 

♦Hevea brasiliensis (Kunth) Miill. Arg. 
♦Hevea guianensis Aubl. 
♦Jatropha curcas L. 
♦Jatropha multifida L. 
Klaineanthus gaboniae Pierre ex Prain 
Macaranga monandra Miill. Arg. 
Macaranga occidentalis (Miill. Arg.) Miill. Arg. 
Macaranga spinosa Miill. Arg. 
Maesobotrya dusenii (Pax) Hutch. 
Maesobotrya glabrata (Hutch.) Exell 

Thecacoris glabrata Hutch. 

Maesobotrya 'sp. no v.' 



16 






A 


E 


I 










I, 2, II, 


12, 


25 


ST A 




12 






FP 


E 


2 










X 






ST 


E 


I 






ST 


E 


I 






ST? 




2, 12 . 






FP 


E 


I, 2, 12 






FPP 




I. 25, 26 






PST 




12 






FP 




I 






P 


E 


2, 12 






FP 


E 


2 










I, II 






ST 


E 


12 






FP 


E 


2 










I, 12 






ST 




7 






FP 




I, 9, II, 


12, 


25. 


FP P ST A 




26 










II, 12 






PST 




I 










I 






ST 




I. 9, II, 


12, 


13. 


FP P ST A 




25, 26 










16 






A 





I 






P 


I 






ST 


I. 2 






FP ST 


I 






ST 


I, 9, II, 


12 




FP ST A 


I, II, 16 


25, 


26 


PST A 


II, 12 






ST 


II, 25 






ST 


2, 12 






FP 


2, 12 






FP 


12 






FP 


15 






P 


I 








I 









354 



ANGIOSPERMS OF THE ISLANDS 



Maesobotrya sp. 

Thecacoris stenopetala sensu Exell pro parte 
Mallotus subulatus Miill. Arg. 
♦Manihot esculenta Crantz 
♦Manihot glaziovii Miill. Arg. 
Manniophyton fulvum Miill. Arg. 

M. africanum Miill. Arg. 
Mareya micrantha (Benth.) Miill. Arg. subsp. 

MICRANTHA 

M. spicata Baillon 
Margaritaria discoidea (Baillon) Webster 

Phyllanthus discoideus (Baillon) Miill. Arg. 
Microdesmis puberula Hook. f. ex Planchon 
Neoboutonia mannii Benth. 
Phyllanthus amarus Schumacher 
Phyllanthus capillaris Schumacher 
Phyllanthus muelleranus (Kuntze) Exell 
Phyllanthus nigericus Brenan 
Phyllanthus odontadenius Miill. Arg. 

P. niruri sensu Exell 
Phyllanthus pentandrus Schumacher 
Phyllanthus physocarpus Miill. Arg. 
Phyllanthus reticulatus Poiret 
Phyllanthus tessmannii Hutch. 
Protomegabaria macrophylla Hutch. 
Pseudagrostistachys africana (Miill. Arg.) Pax & 

Hoffm. subsp. AFRICANA 

Pycnocoma macrophylla Benth. var. macrophylla 

RiCINODENDRON HEUDELOTII Subsp. AFRICANUM (Miill. 

Arg.) J. Leonard 

R. africanum Miill. Arg. 

See J. Leonard in Bull. Soc. Bot. Brux. 31 : 398 (1961). 
♦RiciNUS communis L. 
Sapium ellipticum (Hochst.) Pax 
Spondianthus preussii Engl. var. preussii 
Tetracarpidium conophorum (Miill. Arg.) Hutch. & 

Dalziel 
Tetrorchidium didymostemon (Baillon) Pax & Hoffm. 
Thecacoris annobonae Pax & Hoffm. 
Thecacoris manniana (Miill. Arg. )Miill. Arg. 
Thecacoris membranacea Pax 
Thecacoris stenopetala (Miill. Arg.) Miill. Arg. 
Tragia benthamii Baker 
Tragia tenuifolia Benth. 
♦Tragia volubilis L. 
Uapaca guineensis Miill. Arg. 

Uapaca sp. 
Uapaca staudtii Pax 

66. rosaceae 

Alchemilla cryptantha Steudel ex A. Rich. 

A. tenuicaulis Hook. f. 
♦Cydonia oblonga Miller 

Pyrus cydonia L. 



14. 15 




ST 




I 








2, 12 


FP 






I. 13 




ST A 




I 




PST 




12 




PST 




I, 2 






^ 


12 


FP 




t 


2 




PST 




I, II, 25 








12 


FP 






I, 2, 12 


FPPST 




2, 9, II, 12, 13 


FP 


ST A 




7, 12 


FP 






I, 12 


FP 


ST 




12 


FP 






I, II, 12 


FPPST 




I 








I, 9, II 




A 




I 




P 


E 


16 




A 




I 




ST 




I, 12 




PST 




I, 2, 12 


FP 


ST 




12 


FP 






12 


FP 




1 


7 






* ; 

! 


I, 13. 16 




PST A 


i 


I, 12 


FP 


ST 




12 


FP 




1 


2, 12 


FP 






I, II, 12, 25 


FP 


ST 




I, 16 




A 


E 


I. 13, 15 




ST 


E 


I 




ST 


E 


I (p.p.), 2, 12 


FPP 


1 


2, 12 


FP 






I 




ST 


1 


2, 12 


FP 




1 

■ 


12, 27 


FP 


ST 




25 








2, 12 


FP 






I, 12, 13 


FP 


ST 




2 








13 




ST 




13 










13 




ST? 
ST 


13 






I, II, 13 




ST 


2, 12, 15 


FP 


ST 


I 




ST 


I 




ST 


15 




ST 


2 


FP 





■ OF THE GULF OF GUINEA 355 

H^Eriobotrya japonica (Thunb.) Lindl. i, 13, 15 ST 

■♦Malus pumila Miller i ST ? 

A doubtful record. 
♦Prunus cerasus L. 1 ST ? 

A doubtful record. 
♦Prunus domestica L. 
♦Prunus dulcis (Miller) D. A. Webb 

P. amygdalus Batsch 
♦Prunus persica (L.) Batsch 
Pygeum africanum Hook. f. 

If transferred to Prunus this becomes P. africana 

(Hook, f.) Kalkman 
♦Rosa chinensis Jacq. 
♦Rosa laevigata Michaux 
♦Rosa multiflora Thunb. 
♦RuBUS DISCOLOR Weihe & Nees 
'Rubus extensus Schube ex Engl.' Index Kew., Suppl. i : 

370 (1906), cited as 'in Abh. Preus. Akad. Wiss. 1891 

(1892) 2X2. — Ins. Fernando Po'. A manifest error for 

Ranunculus extensus Schube ex Engl., torn. cit. : 218. 
RuBUS piNNATUs var. AFROTROPicus (Engl.) Gustafsson 

Rubus afrotropicus (Engl.) Guinea 
♦RuBUS ROSiFOLius Smith 
Rubus sp. 

67. CHRYSOBALANACEAE 

AciOA MANNii (Oliver) Engl. 

Chrysobalanus ellipticus Solander ex Sabine 

? Chrysobalanus orbicularis Schumacher 

Probably C. ellipticus. 
Parinari excelsa Sabine 

A doubtful record. 
Parinari macrophylla Sabine 

A doubtful record. 
Parinari mobola Oliver 

A doubtful record. 

68. dichapetalaceae 

DicHAPETALUM AFFiNE Planchon ex Benth. 
DicHAPETALUM BOCAGEANUM (Henriq.) Engl. 
DiCHAPETALUM FLORiBUNDUM (Planchon) Engl. 
DicHAPETALUM kamerunense Engl. 

FP : San Carlos, Mildbraed Oygs. Not in 12. 
DicHAPETALUM oblongum (Hook. f.) Engl. 
DicHAPETALUM subcordatum (Hook. f.) Engl. 

DicHAPETALUM TOMENTOSUM Engl. 

D. acutifolium Engl. 
FP : 'Bdkdko.Burchardt. Not recorded from FP in 12. 
Tapura africana Oliver 2, 12 FP 

69. LEGUMINOSAE 

Abrus precatorius L. I, 9, n. 12 FP ST A 

♦AcACIA FARNESIANA (L.) Willd. I, 2, 25 FP ST 

Acacia kamerunensis Gandoger 15 ST 

A. pennata sensu auctt., pro parte i, 25 



I, 2, 


12, 


13 


FP 


ST A 




2 

I, 8, 

15 


II, 


13 


FPPST 
ST 




2, 7. 
12 

7 


12 




FP 
FP 
FP? 






I, 13 








ST? 




I, 12 








ST? 




13 








ST? 




12 

I 
2, 12 






FP 

FP 
FP 


ST 


E 
E 


2, 12 
2, 12 






FP 
FP 
FP 




E 



356 



ANGIOSPERMS OF THE ISLANDS 



Acacia pentagona (Schumacher) Hook. f. 

A. pennata sensu auctt., pro parte 
♦Adenanthera pavonina L. 
Adenocarpus mannii (Hook, f.) Hook. f. 
Aeschynomene indica L. 

AlRYANTHA SCHWEINFURTHII Subsp. CONFUSA (Hutch. & 

Dalziel) Brummitt 

Baphia confusa Hutch. & Dalziel 

Baphiastrum confusum (Hutch. & Dalziel) Pellegrin 

Baphia spathacea Hook. f. 
See Brummitt in Kew. Bull. 22 : 381 (1968). 
♦Albizia falcata (L.) Backer 
♦Albizia lebbeck (L.) Benth. 
♦Albizia moluccana Miq. 
♦Albizia procera (Roxb.) Benth. 
Albizia zygia (DC.) J. F. Macbride 

A. brownei (Walp.) Oliver 
Not recorded from FP in 12. 

Alysicarpus vaginalis (L.) DC. 
Anthonotha macrophylla Beauv. 

Macrolohium palisotii Benth. 
♦Arachis hypogaea L. 
Baikiaea insignis Benth. subsp. insignis 
Baphia capparidifolia subsp. polygalacea (Hook, f.) 
Brummitt 

B. polygalacea (Hook, f.) Baker 
Baphia nitida Lodd. 

ST : Espirito Santo 4422. 
♦Bauhinia purpurea L. 
♦Bauhinia tenuiflora Watt ex C. B. Clarke 
♦Bauhinia tomentosa L. 
Berlinia bracteosa Benth. 
Caesalpinia bonduc (L.) Roxb. 

C. crista sensu Guinea 
♦Caesalpinia pulcherrima (L.) Swartz 
♦Cajanus cajan (L.) Millsp. 

C. indicus Sprengel 
♦Calopogonium mucunoides Desvaux 
Canavalia ensiformis (L.) DC. 

Not recorded from FP in 12. 
Canavalia maritima (Aubl.) Urban 
C. rosea (Swartz) DC. 
C. ohtusifolia DC. pro parte 
Not recorded from FP in 12. 
♦Cassia alata L. 
Cassia kirkii Oliver 
♦Cassia laevigata Willd. 
Cassia mannii Oliver 
Cassia mimosoides L. 

FP : Moka, Tessmann 2868. Not recorded from FP 
in 12, 
Cassia ohtusifolia L. 
C. tora sensu auctt. 
See Brenan in Kew Bull. 13 : 248 (1958). 



15 






P 


I 








2, 13 




FP 


ST 


2, 7, 12 




FP 




I. 13 




FP 


PST 


2 








12 








2 








II. 13 






ST 


2, II, 12 




FP 


ST 


2 




FP 




II 




FP 


ST 


2 








I, 9, II 






ST A 


12 




FP 




2 








I, 3. 16, 


25 


FP 


ST A 


2, 12 




FP 
FP 




2, 12 








12 




FP 


ST 


I, II 






ST 


2 




FP 




27 






ST 


2, 12 




FP 




I, 9. II. 


12, 25 


FP 


ST A 


7 








I. 2, II, 


12, 16 


FP P ST A 


I, 16, 25 




FP P ST A 


2 








27 






ST 


I. 2 




FP 


ST 


27 




FP 


ST A 


I. 13 








7 








2, 27 




FP 


ST 


12 




FP 




I, 2, II, 


12 


FP 


ST A 


I, II. 12 






PST 


I 




FP 


ST 


16 




FP 


ST A 


I, 12 









OF THE GULF OF GUINEA 



357 



♦Cassia occidentalis L. 

Cassia podocarpa Guillemin & Perrottet 

♦Cassia siamea Lam. 

Cassia sieberana DC. 

♦Cassia sophera L. 

♦Cassia spectabilis DC. 

♦Centrosema plumieri (Turpin ex Pers.) Benth. 

♦Centrosema pubescens Benth. 

♦Cercis siliquastrum L. 

♦Clitorea ternatea L. 

Colophospermum mopane (Benth.) Kirk ex J. Leonard 
Copaifera mopane Benth. 
Helder Lains e Silva (13) records Copaifera mopane 
Cpau ferro') as one of the important timber trees of 
S. Tome, but no one else to my knowledge has ever 
recorded it from the island and I have seen no specimens. 

Crotalaria doniana Baker 
ST : Espirito Santo 66. 

Crotalaria goreensis Guillemin & Perrottet 

♦Crotalaria lanceolata E. Meyer 

Not recorded from FP in 12. If correctly named, not 
indigenous. 

Crotalaria mucronata Desvaux 

Crotalaria ochroleuca G. Don 

Crotalaria retusa L. 

P : Rozeira 432. Not recorded from FP in 12. 

♦Crotalaria zanzibarica Benth. 
C. thomensis Baker f. 

Crudia klainei Pierre ex De Wild. 

C. senegalensis sensu Guinea 
Cynometra mannii Oliver 

Dalbergia ecastaphyllum (L.) Taub. 
♦Dalbergia monetaria L. 
Daniellia oblonga Oliver 

D. thurifera sensu Exell 
Daniellia ogea (Harms) Rolfe ex Holland 
♦Delonix regia (Bojer ex Hook.) Raf. 

Poinciana regia Bojer ex Hook. 
♦Desmanthus virgatus (L.) Willd. 
Desmodium adscendens (Swartz) DC. var. adscendens 
Desmodium adscendens var. robustum Schubert 
Desmodium canum (J. F. Gmelin) Schinz & Thell. 

D. incanum DC. 
Desmodium gangeticum (L.) DC. 

P : Gossweiler s.n., Rose 487, Rozeira 4220. 
Desmodium procumbens (Miller) A. S. Hitchc. 

D. ospriostreblum Steudel ex Chiov. 
Desmodium ramosissimum G. Don 

D. mauritianum sensu Exell 
Desmodium repandum (Vahl) DC. 

D. scalpe DC. 



I, 2, II, 


12, 


16, 


FP P ST A 


25, 26 








I, 2, 12, 


25, 


26 


FP ST 


27 






ST 


I 






ST ? 


I. 2, 25 






FPPST 


27 






ST 


II 






ST 


II 






ST 


13 






ST 


2, II 






FP ST 
ST? 



ST 



12 




FP 


2 




FP 


I, 9, II 




PST A 


I. 12 




ST 


I. 2, 13 




FPPST 


15 




ST 


1, II, 14 






2, 12 




FP 


2 






I, 9, II, 12 


13. 


ST 


25 






I, 2, II, 12 


13 


FP P ST A 


2 




FP 


2, 12 




FP ST ? 


I 






12 




FP 


I 




FP ST 


2 






II, 13, 15. ' 


25 


PST 


I, 2, 12, 13 




FP P ST A 


12 




FP 


I, 8, 9, II, 


12, 


FP P ST A 


13 






2 






X 




P A 


II 




A 


I, 9 






II, 12, 16 




FP P ST A 


I, 13 






II, 12 




FP ST 


I, 2, 13 







358 



ANGIOSPERMS OF THE ISLANDS 



Desmodium tortuosum (Swartz) DC. 

ST : Espirito Santo 4203, Rozeira 50J2. 
Desmodium triflorum (L.) DC. 
? D. trifolium sensu Guinea 
Desmodium velutinum (Willd.) DC. 
D. lasiocarpum (Beauv.) DC. 

D. lati folium DC. 

Not recorded from FP in 12. 
DiALIUM guineense Willd. 
Not recorded from FP in 12. 

DiOCLEA REFLEXA Hook. f. 

Distemonanthus benthamianus Baillon 

Not recorded from FP in 12. 
Drepanocarpus lunatus (L. f.) G. F. W. Meyer 

ST : Rozeira 4403. 
Entada africana Guillemin & Perrottet 

Not recorded from FP in 12. 
Entada gigas (L.) Fawcett & Rendle 

E. scandens sensu Guinea 
Entada mannii (Oliver) Tisserant 

Piptadenia mannii Oliver 
Eriosema glomeratum (Guillemin & Perrottet) Hook, f . 
? Eriosema parviflorum E. Meyer subsp. parviflorum 

A doubtful record. 
? Erythrina caffra Thunb. 
* Erythrina droogmansiana De Wild. & Dur. 

Probably introduced. 
Erythrina gilletii De Wild. 

Not in 12. 
♦Erythrina indica Lam. 
♦Erythrina poeppigiana (Walp.) O. F. Cook 
? Erythrina senegalensis DC. 

This may be E. vogelii. 
♦Erythrina umbrosa Kunth 
♦Erythrina velutina Willd. 
Erythrina vogelii Hook, f . 
Erythrophloeum guineense G. Don 

Not recorded from FP in 12. 
Glycine wightii (Wight & Arn.) Verde. 

G. javanica sensu Exell 
Griffonia physocarpa Baillon 
♦Haematoxylum campechianum L, 
Indigofera astragalina DC. 

ST : Espirito Santo 4214. 
Indigofera colutea (Burm. f.) Merr. var. colutea 

ST : Espirito Santo 4228. 
Indigofera hirsuta L. 
Indigofera spicata Forssk. 

/. hendecaphylla Jacq. 
Indigofera suffruticosa Miller 
Indigofera tinctoria L. 
Indigofera trita var. subulata (Vahl ex Poiret) Ali 

ST : Rozeira 4416. 
♦Inga edulis Mart. 







ST 


13 




FP ? P ST 


2 






II 




FP P ST A 


I, 9. 


13 




2 






I- 7. 


II, 12, 25 


FPPST 


I, 12 


, 25 


FPPST 


2 




FP 


I 




PST 


2 




FP 


12 




FP 


2 






12 




FP 


2 






16 




A 


12 




FP? 


I 
27 




P? 
ST 



FP 



13, 14. 


15 




PST 


15 






ST 


2 




FP 


? 


13 






ST 


I, 13 






ST A 


12 
2 




FP 
FP 


ST 


II 








2, 12 
27 




FP 


ST 



ST 



ST 



II, 13 


ST 
ST A 


I, 9 

II, 25 

I, 9, II, 25 


ST 
PST A 

ST 


15. 25, 27 


ST ■ 



OF THE GULF OF GUINEA 



359 



*Inga sp. 

Lablab purpureus (L.) Sweet 
Dolichos lablab L. 
Not recorded from FP in 12. 
Leptoderris brachyptera (Benth.) Dunn 

L. africana (Baker) Harms 
♦Leucaena leucocephala (Lam.) De Wit 

L. glauca sensu Benth. 
Leucomphalus capparideus Benth. ex Planchon 
Lonchocarpus cyanescens (Schumacher) Benth. 
LoNCHOCARPUs GRiFFONiANUs (BaiUon) Dunn 
? Millettia thonningii sensu Guinea 
ST : Espirito Santo 4553. 
Lonchocarpus sericeus (Poiret) Kunth ex DC. 
Millettia barteri (Benth.) Dunn 
Millettia macrophylla Benth. 
Millettia mannii Baker 

Not recorded from FP in 12. 
Millettia mildbraedii Harms 

FP : San Carlos, Mildbraed 6gy5. Not in 12. 
Millettia sanagana Harms 
Millettia thonningii (Schumacher) Baker 
♦Mimosa polydactyla Humb. & Bonpl. ex Willd. 
♦Mimosa pudica L. 

M. pigra sensu Espirito Santo 
This may be M. polydactyla. 
MucuNA FLAGELLiPES T. Vogel ex Hook. f. & Benth. 
Mucuna pruriens (L.) DC. 
M. utilis sensu Guinea 
Not recorded from FP in 12. 
Mucuna sloanei Fawcett & Rendle 
M. urens sensu Exell 

OrMOCARPUM SENNOIDES Subsp. HISPIDUM (Willd.) 

Brenan & Leonard 

O. cochinchinense (Lour.) Merr. 
See Brenan & Leonard in Bull. Jard. Bot. Brux. 24 : 104 

(1954)- 
Ormocarpum verrucosum Beauv. 
♦Ormosia dasycarpa Jackson 

OSTRYOCARPUS RIPARIUS Hook. f. 

Parkia clappertonia Keay 
? P. africana sensu Guinea 
Not recorded from FP in 12. 
Parkia oliveri J. F. Macbride 
♦Peltophorum pterocarpum (DC.) Backer ex K. Heyne 

P. ferrugineum (Decaisne) Benth. 
Pentaclethra macrophylla Benth. 

Phaseolus adenanthus G. F. W. Meyer 
Phaseolus lunatus L. 
♦Phaseolus vulgaris L. 
Physostigma venenosum Balf. 
? PiLiosTiGMA reticulatum (DC.) Hochst. 
Bauhinia reticulata DC. 
Not recorded from FP in 12. 



13 

2 


ST 

FP 


2, 12 


FP 


2 


FPPST 


1, 2, 12, 13 

2, 12 


FP 


2, 12 
2, 7, 12 


FP 

FP ST 


2 




1, 11, 12, 13 

I 

2, 12 
2 


FPP ST 
ST 
FP 
FP 



I, II 



I, 9. II, 13 



FP 



2, 12 


FP 


I, 12, 25 


FP ST 


I 


PST 


2, 27 


FP ST 


25 




2, 12 


FP 


I. 2, 25 


FP ST 


2 (photo) 




12, 16, 25 


FP ST 


I 





I, 2, 12 


FPPST 


13 


ST 


2, 12 


FP 




FP 



E? 



ST 



ST A 









FP 


2 








I, 2, 7, 


12, 


13, 


FP P ST A 


25 








I 






ST 


I, 2, 12 






FP ST A 


I 






ST 


2, 12 






FP 


27 






FP ? ST ? 


2 









36o ANGIOSPERMS OF THE ISLANDS 

PiPTADENiASTRUM AFRiCANUM (Hook. f.) Brenan FP 

Piptadenia africana Hook, f . 
FP : Bokoko, Burchardt. Not recorded from FP in 12. 
PsoPHOCARPUs scANDENS (Endl.) Verdc. 27 ST 

♦Pterocarpus indicus Willd. x ST 

ST : Espirito Santo 4388. 
Pterocarpus mildbraedii Harms 2, 12 FP 

♦PuERARiA jAVANiCA Benth. 13 ST 

Rhynchosia densiflora (Roth) DC. 12 ST 

R. debilis G. Don i 

Rhynchosia mannii Baker 2, 12 FP 

Rhynchosia minima (L.) DC. ST A 

R. memnonia sensu Exell i, 9, 11 

Rhynchosia preussii (Harms) Taub. ex Harms 12 FP 

R. cyanosperma Benth. ex Baker pro parte 2 

Rhynchosia pycnostachya (DC.) Meikle 12 FP 

Samanea saman (Jacq.) Merr. FP ST 

Pithecellobium samana (Jacq.) Benth. 2, 13 

Scorodophloeus zenkeri Harms 2 FP 

Not in 12. 
Sesbania sericea (Willd.) Link ST A 

S. pubescens DC. i, 9, 11, 12 

SOPHORA tomentosa L. ST 

S. occidentalis L. i, 12 

*Tamarindus indica L. I, 2, 13, 25, 26 FP P ST A 

♦Tephrosia CANDIDA (Roxb.) DC. II, 25 ST 

Tephrosia noctiflora Bojer ex Baker ST 

ST. : Espirito Santo 4324, Rozeira yy^, 1231. 
Tephrosia platycarpa Guillemin & Perrottet ST 

T.flexuosa G. Don i, 12 

Tephrosia purpurea subsp. leptostachya var. 

pubescens Baker 
Tephrosia uniflora Pers. 
ST : Espirito Santo 4636. 
Tephrosia vogelii Hook. f. 

Teramnus labialis (L. f.) Sprengel 
Tetrapleura tetraptera (Schumacher) Taub. 
Trifolium rueppellianum var. preussii (Taub. ex 
Baker f.) J. B. Gillett 

T. preussii Taub. ex Baker f. 
Trifolium simense Fresen. 

T. subrotundum Steudel ex Hochst. 
Trifolium usambarense Taub. 

T. mannii Guinea, nom. nud. 
Uraria picta (Jacq.) DC. 
*VlCIA faba L. 

ViGNA alba (G. Don) Planchon ex Baker f . 
Vigna gracilis (Guillemin & Perrottet) Hook. f. 

V. occidentalis Baker f . 

ViGNA MULTIFLORA Hook. f. 

ViGNA OBLONGA Benth. 

V. marina sensu Exell 



I 








ST 
ST 


I. 2, 


II, 


12, 


25. 


FPPST 


26 










I. 9. 


II, 


15 




PST 


I, 12 


13 






ST 


12 








FP 


2 










2, 12 








FP 


2 










12 








FP 


7 










I, 9, 


II 






ST 


13 








ST 


I, 13 








ST 


12 








P 


I 










2, 12 








FP 


2 








FP ST 


I. 12 











OF THE GULF OF GUINEA 361 

ViGNA RACEMOSA (G. Don) Hutch. & Dalziel i, 8, 11. 13 P ST 

P : Rozeira 3932, 3ggi. 
ViGNA UNGUICULATA (L.) Walp. i6 FP A 

V. sinensis Endl. ex Hassk. 
FP : Mann 415. Not recorded from FP in 12. See 
Baker in Oliver, Fl. Trop. Afr. 2 : 205 (1871). 
ViGNA VExiLLATA (L.) A. Rich. 12 FP 

V. thonningii Hook. f. 2 

Zenkerella citrina Taub. 12 FP 

Cynometra leptoclada Harms 2 

Zenkerella pauciflora Harms FP 

Cynometra pauciflora (Harms) Harms 2 

Not in 12, 



♦BUXUS SEMPERVIRENS L. I ST ? 



Myrica arborea Hutch. 12 FP 



70. BUXACEAE 






I 


71. MYRICACEAE 






12 


72. ULMACEAE 






I, II. 12, 13, 25, 




27 




13 




12, 25, 27 




I. 13 




I, 9, II, 27 




2 



Celtis gomphophylla Baker ST 

C. durandii Engl. 

Celtis integrifolia Lam. 13 ST ? 

Celtis mildbraedii Engl. 12, 25, 27 ST 

C. soyauxii sensu G. Taylor 
Celtis prantlii Priemer ex Engl. i, 9, 11, 27 P ST A 

C. insularis Rendle 
This is considered by Letouzey [Fl. Cameroun 8 : 26 
(1968)) to be C. philippensis Blanco. 
Celtis wrightii Planchon FP 

C brownii Rendle 12 

C. prantlii sensu Guinea 2 

C. rendleana G. Taylor i, n 

See Polhill in Kew Bull. 19 : 141 (1964). Letouzey 
{loc. cit.) considers that this also is C. philippensis 
Blanco. 
Chaetacme aristata E. Meyer ex Planchon i, 12 A 

Trema orientale (L.) Blume FP P ST A 

T. guineense (Schumacher) Ficalho i, 2, 7, 12, 13 

73. moraceae 
Antiaris africana Engl. 2 FP 

Not recorded from FP in 12. 
Antiaris sp. FP 

FP : Bokoko, Mildhraed 6880. 
♦Artocarpus altilis (S. Parkinson) Fosberg FP P ST A 

A. communis J. R. & G. Forster i, 13, 16, 25 

A . incisus (Thunb.) L. f , 2 

*Artocarpus heterophyllus Lam. 16, 25 FP P ST A 

A. integer sensu G. Taylor i, 13 

A. integrifolius sensu Guinea 2 



362 



ANGIOSPERMS OF THE ISLANDS 



BOSQUEIA ANGOLENSIS Ficalho 

♦Castilloa elastica Cerv. 

♦Cecropia peltata L. 

Chlorophora excelsa (Welw.) Benth. 

DoRSTENiA barteri Bur. 

DORSTENIA ELLIPTICA BuF. 
DORSTENIA PROREPENS Engl. 

Ficus ANNOBONENSis Mildbr. & Hutch. 

Not in 12. Recorded from FP in i, fide G. Taylor. 
Ficus camptoneura Mildbr. 
Ficus camptoneuroides Hutch. 
Ficus capensis Thunb. 
♦Ficus carica L. 
Ficus chlamydocarpa Mildbr. & Burret 

jF. clarencensis Mildbr. & Hutch. 
♦Ficus elastica Roxb. 
Ficus eriobotryoides Kunth & Bouche var. 

eriobotryoides 
Ficus exasperata Vahl 
Ficus fernandesiana Hutch. 
Ficus globicarpa Warb. ex Guinea, nom. nud. 

I cannot trace that this name, listed by Guinea in 2, 

has been validly published. 
Ficus kamerunensis Warb. ex Mildbr. & Burret 
Ficus macrosperma Warb. ex Mildbr. & Burret 

FP : Melville (&> Wrigley 622 (K). Not recorded from 

FP in 12. 
Ficus mallotocarpa Warb. 
Ficus mucuso Welw. ex Ficalho 
Ficus natalensis Hochst. 
Ficus ottoniifolia (Miq.) Miq. 
Ficus ovata Vahl 
Ficus praticola Mildbr. & Hutch. 
Ficus preussii Warb. 

Not recorded from FP in 12. 
♦Ficus pumila L. 
'Ficus scabra (var. Jacq. nee Wild.)' 

I do not know what this is. 
Ficus sidifolia Welw. ex Hiern 
Ficus spirocaulis Mildbr., nom. nud. 
? F. urceolaris sensu Guinea 

FP : Musola, Mildbr aed yoio. 
Ficus thonningii Blume 

F. microcarpa Vahl 
Ficus vogeliana (Miq.) Miq. 
Ficus vogelii (Miq.) Miq. 

F. vogelii var. pubicarpa Mildbr. & Burret 
Ficus wildemaniana Warb. 

Not in 12. 
Mesogyne henriquesii Engl. 
MUSANGA cecropioides R.Br. 

M. smithii R.Br. 



I, 12 


A 


I, II 


PST 


11. 25 


ST 


I, 2, 7, II, 12, 


FPPST 


13, 25 




2, 12 


FP 


2, 12 


FP 


2, 12 


FP 


I, 2, II 


FP P ST A 


2, 12 


FP 


2, 12 


FP 


I, 25 


ST 


13 


ST 


12 


FP P ST A 


I. 2, 13 




I 


ST 


12 


FP 


I, 2, 7, 12, 25 


FP ST 


II 


ST 


2 


FP 



1, 12, 25 



FP 



ST 



2, 12 


J 


FP 






I, 12 




FP 


ST 




2, 12 




FP 






2, 12 




FP 






2, 7. 


12 


FP 






2, 12 




FP 






2 




FP 






27 






ST 




2 




FP 






9, II, 


13 




ST A 




2, 12 




FP 




E? 


2. 7 










12 




FP 






2 










2, 7. 


12 


FP 






I, 12 




FP 


ST 




2 

2 




FP 






I, II 






ST 


E 


I. 7. 


II, 12, 13, 


FP 


ST 




25 










2 











OF THE GULF OF GUINEA 



363 



Myrianthus preussii Engl. 

Treculia africana Decaisne var. africana 

Not recorded from FP in 12. 



12 


FP 


I, 7, II, 12, 13, 


FPPST 


25 





74. URTICACEAE 

♦BoEHMERiA NiVEA (L.) Gaudich. 

BOEHMERIA PLATYPHYLLA D. Don 

Droguetia iners (Forssk.) Schweinf. 

Not recorded from FP in 12. 
Elatostema mannii Wedd. 
Elatostema monticola Hook. f. 
Elatostema paivaeanum Wedd. 
Elatostema thomense Henriq. 

Perhaps a form of E. paivaeanum. 
Elatostema welwitschii Engl. 

E. henriquesii Engl. 
Laportea aestuans (L.) Chew 

Fleurya aestuans (L.) Gaudich. ex Miq. 

Laportea alatipes Hook. f. 

Not recorded from FP in 12. 
Laportea ovalifolia (Schumacher) Chew 
Fleurya ovalifolia (Schumacher) Dandy 

F. podocarpa Wedd. 

Lecanthus peduncularis (Royle) Wedd. 
Parietaria debilis Forster f. 

P. laxiflora Engl. 
PiLEA manniana Wedd. 
PiLEA microphylla (L.) Liebm. 
PiLEA RivuLARis Wedd. 

p. ceratomera Wedd. 
PiLEA SUBLUCENS Wedd. 
PiLEA TETRAPHYLLA (Steudel) Blume 
PouzoLziA GUiNEENSis Benth. 

PouzoLziA PARASITICA (Forssk.) Schwcinf. 

Not recorded from FP in 12. 
Procris crenata C. B. Robinson 

P. laevigata sensu Guinea 
Urera batesii Rendle 
Urera cameroonensis Wedd. 
Urera mannii (Wedd.) Benth. & Hook, ex Rendle 
Urera repens (Wedd.) Rendle 
Urera thonneri De Wild, ex T. Dur. 

FP : Escarrd 451. Not in 12. 



I, 2, 13 


FP 


ST 


I, 2, 12, 19 


FP 


ST 


19 


FP 




2, 12, 19 


FP 




12, 19 


FP 




2, 12, 19 


FP 




I. 13 




ST 


I, 2, 7, II, 12, 


FP 


ST 


13, 19 






2 








FP P ST A 


I, 2, II, 12, 13, 






19, 26 






19 


FP 
FP 




12, 19 






2 






2, 12, 19 


FP 




19 


FP 




2, 12 






I, 13 




ST 


2, 12, 19 


FP 




19 


FP 


ST 


I, 2, 12, 13 






2, 12, 19 


FP 




12, 19 


FP 




I, 2, II, 12, 13, 


FP 


ST 


19 






19 


FP 




I, 12, 19 


FP P ST A 


2 






2, 12, 19 


FP 




I, 2, 12, 16, 19 


FP 


ST A 


I, 2, 12, 19 


FP P ST A 


2, 12, 19 


FP 




19 


FP 





75. cannabaceae 



♦Cannabis sativa L. 



I. 13 



ST 



76. AQUIFOLIACEAE 



Ilex mitis (L.) Radlk. 



12, 27 



FP ST 



12 


FP 




12 


FP 




2 






I, 12 




ST 


16 




A 


I, 2 









2, 12 

2, 12 


FP 
FP 




I, 12 




ICACINACEAE 








2, 12 
2, 12 
12 


FP 
FP 
FP 


& Mendon9a 


2 
12 


FP 
FP 



364 ANGIOSPERMS OF THE ISLANDS 

77. CELASTRACEAE (incl. HIPPOCRATEACEAE) 

Campylostemon angolense Welw. ex Oliver 

HiPPOCRATEA INDICA Willd. 

H. loesenerana Hutch. & Moss 

HiPPOCRATEA VELUTINA Afzel. 

Maytenus annobonensis (Loes. & Mildbr.) Exell 

Gymnosporia annobonensis Loes. & Mildbr. 
Maytenus buchananii (Loes.) R. Wilczek FP 

FP : Wrigley <S' Melville 673 (K.) Not in 12. 
Maytenus monodii Exell 15 ST 

Salacia cf. cuspiDicoMA Loes. FP 

FP : Bokoko, Burchardt. 
Salacia debilis (G. Don) Walp. 
Salacia mannii Oliver 
Salacia pyriformis (Sabine) Steudel i, 12 ST 

78 

Desmostachys tenuifolius Oliver 
Lasianthera africana Beauv. 
Lavigeria macrocarpa (Oliver) Pierre 

Icacina macrocarpa Oliver 
Neostachyanthus zenkeri (Engl.) Exell & Mendon9a 
Rhaphiostylis ferruginea Engl. 

FP : Bokoko, Burchardt. Not recorded from FP in 12 

79. OLACACEAE 

Heisteria parvifolia Smith 
Olax gambecola Baillon 

O. viridis Oliver 
Strombosia grandifolia Hook. f. ex Benth. 
Strombosia scheffleri Engl. 
Strombosia sp. 

XiMENIA AMERICANA L. 1, 25, 26 A 

80. OPILIACEAE 

Urobotrya minutiflora Stapf FP 

Opilia minutiflora (Stapf) Engl. 2 I 

Not recorded from FP in 12. 

81. loranthaceae 

General reference : S. Balle, 'Les Loranthacees d'Afrique portugaise. I. Guinee et San Tome' 

in Bolm. Soc. Broter., ser. 2, 38 : 9-80, t. 1-16 (1964). 
? Globimetula braunii (Engl.) Tieghem FP ? 

Loranthus zenkeri Engl. 2 

Not recorded from FP in 12. Balle considers it 
unlikely that the species occurs in FP. 
Helixanthera mannii (Oliver) Danser var. mannii ii, 12, 13 ST 

Loranthus mannii Oliver I 

Loranthus ternatus Tieghem i 

Helixanthera mannii var. ternata (Tieghem) Balle 11 

Phragmanthera capitata (Sprengel) Balle FP 

Loranthus capitatus (Sprengel) Engl. 2 

Loranthus incanus Schumacher 2 

Phragmanthera incana (Schumacher) Balle 12 



I. 7. 12 


FP P ST A 


II 


ST A 


I, 9 




2, 12 


FP 


12 


FP 


I 


A 



OF THE GULF OF GUINEA 



365 



Phragmanthera nitidula (Sprague) Balle 

Loranthus nitidulus Sprague 2 

Phragmanthera polyscripta Balle pro parte 12 

Tapinanthus bangwensis (Engl. & Krause) Danser 12 

ViscuM ENGLERi Tieghem 16 

ViSCUM TRIFLORUM DC. 

V. nervosum var. angustifolium Sprague 1 



FP 



FP 



ST 



Thesium tenuissimum Hook. f. 



82. SANTALACEAE 



FP 



83. RHAMNACEAE 

GouANiA LONGiPETALA Hemsley 12 FP 

Lasiodiscus mannii Hook. f. I, 12 P 

Lasiodiscus milder aedii Engl. i, 12 ST 

Lasiodiscus rozeirae Exell 14 ST 

Maesopsis eminii Engl. 2, 12, 27 FP ST 

Ventilago diffusa (G. Don) Exell i, 11, 12 ST 

ZiziPHus ABYSSiNiCA Hochst. ex A. Rich. 25 ST 

♦ZiziPHUs MAURiTiANA Lam. 13 ST 



84. vitaceae 

Ampelocissus cavicaulis (Baker) Planchon 

Not recorded from FP in 12. 
Ampelocissus macrocirrha Gilg & Brandt 

Stated in 1 2 to be doubtfully distinct from A . cavicaulis. 
Cissus ADENOPODA Sprague 

FP : Melville (syWrigley 435b (K). Not recorded from 

FP in 12. 
Cissus aralioides (Welw. ex Baker) Planchon 
Cissus arguta Hook. f. 

Recorded from FP in Fl. W. Trop. Afr., ed. i, i : 475 

(1928) but not in 12. The FP plant may be C. producta. 
Cissus barbeyana De Wild. & Dur. 
Cissus barteri (Baker) Planchon 
Cissus curvipoda (Baker) Planchon 
Cissus debilis (Baker) Planchon 
Cissus diffusiflora (Baker) Planchon 
Cissus glaucophylla Hook, f . 
Cissus gracilis Guillemin & Perrottet 
Cissus lageniflora Gilg & Brandt 
Cissus mannii (Baker) Planchon 
Cissus oliveri (Engl.) Gilg 
Cissus aff. petiolata Hook. f. 
Cissus producta Afzel. 
Cissus vogelii Hook. f. 
♦VlTIS vinifera L. 



85. LEEACEAE 



Leea guineensis G. Don 
Leea tinctoria Lindl. ex Baker 



2 




FP 




2, 12 




FP 
FP 




I 






P? ST 


2 




FP 


? 


I, 12 






PST 


2, 12 




FP 




I 






ST 


I. 2, 


12, 13 


FP 


ST 


2, 12 




FP 




2, 12 




FP 




2, 12 




FP 




12 




FP 




12 




FP 




I 






ST 


I 






ST 


I. 2, 


12 


FP 


ST 


2, 12 




FP 




I. 13 






ST 


1, 2, 


12 


FP 


ST 


I, 13. 


25 




ST 



366 



ANGIOSPERMS OF THE ISLANDS 



86. RUTACEAE 

Afraegle mildbraedii Engl. 

♦Citrus aurantifolia (Christm.) Swingle 

•Citrus aurantium L. 

♦Citrus grandis (L.) Osb. 

♦Citrus limon (L.) Burm. f. 

C. limonum Risso 
♦Citrus nobilis Lour. 

Clausena anisata (Willd.) Hook. f. ex Benth. 
Fagara annobonensis Mildbr. 
Fagara macrophylla Engl. 

Not recorded from FP in 12. 
Fagara rubescens (Planchon ex Hook, f.) Engl. 

F. melanacantha (Planchon ex Oliver) Engl. 

F. welwitschii Engl. 
Fagara thomensis Engl. 

GlYCOSMIS ? AFRICANA Hook. f. 
♦RUTA CHALEPENSIS L. 

Teclea verdoorniana Exell & Mendon9a 
T. grandifolia sensu Guinea 

87. SIMAROUBACEAE 

Brucea guineensis G. Don 

B. paniculata sensu Mildbr. in Notizhl. Bot. Gart. 
Berl. 12 : 185 (1934) is almost certainly B. 
guineensis. 

Not recorded from FP in 12. 
Hannoa klaineana Pierre & Engl. 

FP : Bokoko, Burchardt. Not recorded from FP in 12. 
♦Quassia amara L. 

88. irvingiaceae 

Desbordesia glaucescens (Engl.) Tieghem 

Not recorded from FP in 12. 
Irvingia gabonensis (Aubry-Lecomte ex O'Rorke) 

Baillon 

89. burseraceae 

Canarium schweinfurthii Engl. 

Not recorded from FP in 12. 
Canarium velutinum Gilg 

Not in 12. 
Dacryodes edulis (G. Don) H. J. Lam 
Pachylohus edulis G. Don 
P. edulis var. mubafo Engl. 
Not recorded from FP in 12. 
Santiria trimera (Oliv.) Aubrev. 

Santiriopsis trimera (Oliv.) Guillaumin ex Engl. 

90. meliaceae 

Carapa procera DC. 
♦Cedrela odorata L. 

C. toona sensu Exell 
Entandrophragma angolense (Welw.) CDC. 



2, 12 
I. 16 

1, 2, 13 

2, 13 



7. 12 
2 



12 

I. 13 
I, 13 

I 
I 

12 

2 



13 



I. 12 



FP 

P A 
FP ST A 
FP ST 
FP 



ST 



FP 



I, 7, 12, 13, 25 FP P ST 



ST 



ST 
ST 
ST? 



FP 



FP 



FP 



FP 



ST 



FP 



FP 



12, 25, 27 


FPPST 


I, II. 13 




2 




25, 26, 27 


PST 


I 




I, 2, 12, 25 


FP ST 


13 


ST 


II 




X2 


FP 



E 



OF THE GULF OF GUINEA 



367 



GuAREA GLOMERULATA Harms 
Heckeldora staudtii (Harms) Staner 
Guarea staudtii Harms 

Not recorded from FP in 12. 
♦Melia azedarach L. 
♦TooNA ciLiATA M. Rocm, 

Cedrela toona Roxb. ex Willd. 
Trichilia grandifolia Oliver 
Trichilia macrophylla Oliver ex Guinea, nom. nud., 

non T. macrophylla Benth. 

Apparently an erroneous citation. 
Trichilia monadelpha (Thonn.) J. De Wilde 

T. heudelotii Planchon ex Oliver 
Trichilia rubescens Oliver 
Turraea glomeruliflora Harms 
TuRRAEA voGELii Hook. f. ex Benth. 



2, 12 


FP 
FP 


2 




I. 2, 25 

25. 27 

I 
I 


FPPST 
ST 

ST 


2 


FP? 




FP 


2, 12 

2, 12 

I, 2, 16 

I, 2, 9, II, 12, 


FP 
FPPST 


13 





91. SAPINDACEAE 

Allophylus africanus Beauv. 

FP : Bokoko, Mildhraed 6yg2 ; Musola, Mildbraed 

yoi8. Not recorded from FP in 12. 
Allophylus grandifolius (Baker) Radlk. 
Allophylus hirtellus (Hook, f.) Radlk. 
Allophylus oreodryadum Gilg ex Mildbr., nom. nud. 

FP : Mildhraed 6406. 
Blighia sapida C. Koenig 
Cardiospermum corindum L. 
Cardiospermum halicacabum L. 
Chytranthus mannii Hook. f. 
Deinbollia insignis Hook. f. 
Eriocoelum kerstingii Gilg ex Engl. 

Not recorded from FP in 12. 
Eriocoelum macrocarpum Gilg ex Engl. 

FP : Bokoko, Bur char dt. Not recorded from FP in 12. 
Eriocoelum sp. 

Laccodiscus ferrugineus (Baker) Radlk. 
Lychnodiscus reticulatus Radlk. 
Paullinia pinnata L. 
♦Sapindus saponaria L. 



I, II, 13, 25, 26 FP ST 



I, 


11 






P ST 


2, 


12 






FP 


2, 


12 






FP 


I 








PST 


I 










I, 


2, 


12 




FP ST 


I, 


II, 


13. 


25 


PST 


2, 


12 






FP 


2 








FP 



FP 



12 


FP 


2, 12 


FP 


2, 12 


FP 


I, 2, 9, 12 


FPPST 


I 





E 



92. melianthaceae 

Bersama abyssinica subsp. PAULLiNioiDES (Planchon) 
Verde. 

B. paullinioides (Planchon) Baker 



12 

2 



FP 



93. ANACARDIACEAE 

♦Anacardium OCCIDENTALE L. 
Antrocaryon klaineanum Pierre 
Lannea welwitschii (Hiern) Engl. 
♦Mangifera indica L. 
PsEUDOspoNDiAS MicROCARPA (A. Rich.) Engl. 



1, 2, 25 

2, 12 

I 

I, 2, 16 

I, 2, II, 12, 13, 
25 



FP P ST A 
FP 

A 
FP P ST A 
FP P ST A 



368 ANGIOSPERMS OF THE ISLANDS 

SORINDEIA GRANDIFOLIA Engl. I, 12, 1 3 ST 

? SORINDEIA JUGLANDIFOLIA Planchon 2 ST ? 

Record probably based on a misidentification of S. 

grandi folia. 
♦Spondias cytherea Sonnerat i, ii, 13, 25 P ST 

S. dulcis Forster f , 2 

♦Spondias mombin L. i, ii, 12, 25, 26 P ST A 

Trichoscypha bijuga Engl. 2, 12 FP 

94. connaraceae 
Agelaea annobonensis Schellenb. i, 2 A E 
Agelaea coccinea Exell i P E 
Agelaea obliqua var. cordata (Schellenb.) Exell i, 12 ST 
Agelaea ovalis Schellenb. i, 2, 16 A E 
Agelaea phaeocarpa Exell i P E 
Agelaea pilosa Schellenb. FP 

FP: Melville &• Wrigley 638 (K). Not recorded 

from FP in 12. 
Agelaea principensis Exell i P E 

Agelaea pseudobliqua Schellenb. 2, 12 FP 

Agelaea reticulata Exell i P E 

Byrsocarpus dinklagei (Gilg) Schellenb. ex Hutch. & 

Dalziel 2, 12 FP 

Cnestis ferruginea DC. ^,12 FP P ST 

FP : Veitch (fide Schellenb. in Engl., Pflanzenr. IV, 

127 : 33 (1938))- 

Not recorded from FP in 12. 
? Cnestis urens Gilg 2 FP ? 

Not in 12. 
CoNNARus AFRicANUS Lam. I ST 

CONNARUS GRIFFONIANUS BaiUon I, 2, 12 FP P 

Paxia zenkeri Schellenb. 2 FP 

Not in 12 and not recorded from FP by Schellenberg. 

95. alangiaceae 
Alangium chinense (Lour.) Harms 12, 27 FP ST 

A. hegoniifolium (Roxb.) Baillon 2 

96. araliaceae 
♦Hedera helix L. I, 3, 15 ST 

PoLYSCiAs FULVA (Hiern) Harms 2, 7, 12, 19 FP 

♦PcLYSciAS GuiLFOYLEi var. LACiNiATA L. H. Bailey 16 A 

PoLYSCiAS QuiNTASii Exell I, II, 13, 15, 25 P ST E 

Schefflera barteri (Seem.) Harms var. barteri i, 12, 19 FP P ST 

S. henriquesiana Harms ex Henriq. i, 11 

Tennant (in Kew Bull. 15 : 332 (1961)) considers 
that this is a synonym of S. barteri var. barteri. 
Not recorded from FP in 12. 
Schefflera hierniana Harms 12, 19 FP 

S. ledermannii Harms 2, 7 

Tennant {loc. cit.) refers this with doubt to S. 
barteri var. barteri ; he does not mention 5. 
hierniana, to which it is referred in 12. 
Schefflera mannii (Hook, f.) Harms i, 2, 7, 12, 13, FP ST A 

14. 19 



OF THE GULF OF GUINEA 



369 



97. UMBELLIFERAE 

Apium leptophyllum (Pers.) F. v. Muell. ex Benth. 
Caucalis melanantha (Hochst.) Benth. & Hook, ex 
Hiern 

C. gracilis sensu Guinea 
Centella asiatica (L.) Urban 

Cryptotaenia africana (Hook, f.) Drude 
♦Eryngium foetidum L. 

♦Foeniculum vulgare Miller 
Hydrocotyle bonariensis Lam. 
Hydrocotyle mannii Hook. f. 
Hydrocotyle sibthorpioides Lam. 

H. monticola Hook, f . 
♦Petroselinum crispum (Miller) A. W. Hill 

P. sativum Hoffm. 

P. hortense auctt. 
Peucedanum winkleri H. Wolff 

PiMPINELLA OREOPHILA Hook. f . 

p. petitianum sensu Guinea 
Sanicula elata Buch.-Ham. ex D. Don 

5. europaea var. elata (Buch.-Ham. ex D. Don) H. 

Wolff 
S. europaea sensu Guinea 



I 


P 


2, 7, 12, 19 


FP 


2 




I, II, 12, 19, 25, 


FPPST 


26 




12, 19 


FP 


I, 2, 8, II, 12, 


FPPST 


13. 19. 25 




2 


FP 


I. 13. 19 


FPPST 


I, 2, 12, 19 


FP ST 


12, 19 


FP 


2 






FP ST 


2 




13 




12, 19 


FP 


2, 7, 12, 19 


FP 


2 




12, 19 


FP 


2 




7 





98. THEACEAE 



Balthasaria mannii (Oliver) Verde. 
Adinandra mannii Oliver 
Melchiora mannii (Oliver) Kobuski 

♦Camellia japonica L. 

♦Camellia sinensis (L.) Kuntze 
Thea sinensis L. 



I. 13 

I 



ST 



ST 
ST 



99. ERICACEAE 

Agauria salicifolia (Commerson) Hook. f. ex Oliver 
Blaeria mannii (Engl.) Engl. 

B. dicranotricha Mildbr. 
Philippia mannii (Hook, f.) Aim & Fries 
Philippia thomensis Henriq. 
♦Rhododendron mucronatum (Blume) G. Don 



2, 7. 17 
2, 7. 17 

2. 7. 17 
I. 13, 15 
I 



FP 
FP 

FP 



ST 
ST 



100. ebenaceae 

? DiosPYROS ELLiOTii (Hiern) F. White 
Maba mannii Hiern 
Not recorded from FP in 17. According to F. White 
almost certainly D. piscatoria. 
DiOSPYROS FERREA (WiUd.) Bakh. 

Maba buxifolia (Rottb.) Pers., non Diospyros 
buxifolia (Blume) Hiern 
DiOSPYROS MELOCARPA F. White 



I 

17 



FP 



FP 



ST 



370 ANGIOSPERMS OF THE ISLANDS 

DiosPYROS piscATORiA Gurke FP 

Maha bipindensis Giirke 
FP : Bokoko, Mildbraed 6950. Not recorded from FP in 

17- 



lOI. SAPOTACEAE 

Chrysophyllum albidum G. Don i, 13, 27 P ST 

Gambeya albida (G. Don) Aubrev. & Pellegrin 25 

Chrysophyllum calophyllum Exell i P 

Chrysophyllum delevoyi De Wild. 17. 27 FP ST 

C. africanum sensu Baker in Oliver, Fl. Trop. Afr. 
3 : 500 (1877) et auctt. (? C. africanum A. DC). 
It is impossible to decide to what species the name 
Chrysophyllum africanum A. DC. applies ; see J. H. 
Hemsley in Kew Bull. 20 : 461-465 (1966). 

Chrysophyllum henriquesii Engl. i P 

Chrysophyllum prunifolium Baker 2, 17 FP 

Chrysophyllum sp. 15 ST 

Manilkara multinervis (Baker) Dubard 25, 27 ST 

MiMusops cf. LACERA Baker i 

Omphalocarpum elatum Miers 7. ^7 FP 

ViNCENTELLA DENSiFLORA (Baker) Pierre i ST 

Vincentella revoluta (Baker) Pierre 17 FP 

Bakerisideroxylon revolutum (Baker) Engl. 2 

102. myrsinaceae 

Afrardisia cymosa Mez i. 17 ^^ 

Afrardisia staudtii (Gilg) Mez FP 

FP : Musola, Mildbraed yoos (as A . haemanthus (Gilg) 

Mez). The genus is not recorded from FP in 17. 
Maesa borjaeana Henriq. FP ST 

M. kamerunensis Mez 7. ^7 

M. lanceolata sensu Exell i, 2, 11, 13 

Maesa lanceolata Forssk. 17 FP 

Rapanea neurophylla (Gilg) Mez 17 FP 

Rapanea thomensis Exell 1,13 ST 

103. loganiaceae 

Anthocleista microphylla Wernham 17 FP P ST 

A. macrocalyx Philipson i, n. 13 

A. micrantha Gilg & Mildbr. ex Hutch. & Dalziel 2 

A. stenantha Philipson i. 14 

Anthocleista scandens Hook. f. i, 2, 13, 17 FP ST 

A. exelliana Monod 13. ^5 

A. sp. nov. I 

NuxiA coNGESTA R.Br, ex Fresen. var. congesta 16, 17 FP 

Lachnopylis annobonensis Mildbr. I 

Nuxia annobonensis Mildbr., nom. nud. 2 

NuxiA CONGESTA var. thomensis (Philipson) J. Lewis 16 ST 

Lachnopylis thomensis Philipson i. 15 

Strychnos ACULEATA Solereder 2,17 FP 

Strychnos isabelina Gilg 2 FP 

FP : Mildbraed 6353. Not in 17. 



OF THE GULF OF GUINEA 



371 



Strychnos quadrangularis Mildbr., nom. nud., non S. 
quadrangularis A. W. Hill 
FP : Bokoko, Mildhraed 6817. Not in 17. 



FP 



104. OLEACEAE 

Jasminum fluminense Veil. 

/. mauritianum Bojer ex DC. 
Jasminum thomense Exell 
♦Olea europaea L. 
Olea hochstetteri Baker 
Olea welwitschii (Knobl.) Gilg & Schellenb. 

Steganthus welwitschii (Knobl.) Knobl. 

105. apocyanaceae 

Alafia multiflora (Stapf) Stapf 

Holalafia multiflora Stapf 
♦Allamanda cathartica L. 
Baissea multiflora a. DC. 

Not recorded from FP in 17. 
♦Catharanthus roseus (L.) G. Don 

Lochnera rosea (L.) Reichenb. f. 
♦Ervatamia divaricata (L.) Burkill 

E. coronaria (Jacq.) Stapf 
FuNTUMiA africana (Benth.) Stapf 

Kickxia africana Benth. 
FuNTUMiA ELASTiCA (Preuss) Stapf 

Kickxia elastica Preuss 
Not recorded from FP in 17. 
♦Landolphia dawei Stapf 
*Nerium oleander L. 
*Plumeria alba L. 
♦Plumeria rubra L. 
Rauvolfia dichotoma K. Schum. 
Rauvolfia macrophylla stapf 
Rauvolfia vomitoria Afzel. 

Not recorded from FP in 17. 
Strophanthus bullenianus Mast. 
Strophanthus preussii Engl. & Pax 
Tabernaemontana brachyantha Stapf 

Conopharyngia brachyantha (Stapf) Stapf 
Tabernaemontana eglandulosa Stapf 

Gabunia eglandulosa (Stapf) Stapf 
Tabernaemontana pachysiphon Stapf 

Conopharyngia pachysiphon (Stapf) Stapf 
Tabernaemontana stenosiphon Stapf 

Conopharyngia stenosiphon (Stapf) Stapf 
Tabernaemontana sp. aff. stenosiphon 

Conopharyngia sp. nov. ? 
Tabernaemontana sp. ? 

Conopharyngia sp. 
*Thevetia neriifolia Juss. 
Voacanga africana Stapf 
Voacanga lemosii Philipson 



II 




ST 


I 






II 




ST 


13 




ST 


15. 26 




ST 


17. 25, 27 




ST A 


1, 13 






17 


FP 




2 






16 




A 


2 


FP 




I, 16 


FP 


ST A 


2 






15 




P 


I 






I, 7, II, 17, 25 


FPPST 


2 






13 


FP 


ST 


2 






I 




ST 


I, 2 


FP 


ST 


2 


FP 




I 


FP 


ST A 


I, II, 13, 25 




ST 


I. 11, 13, 17, 25 




ST 


I, 7, 9, II, 13, 17 


FP P ST A 


2, 17 


FP 




17 


FP 




2, 7. 17 


FP 




2 






17 


FP 




2 




ST 


I. 13. 25 




ST 


I. 13 




P 


I, 15 




ST 


I 






2 






I, 2, 17, 25 


FP 


ST 


I 




ST 



372 ANGIOSPERMS OF THE ISLANDS 

I06. ASCLEPIADACEAE 

♦ASCLEPIAS CURASSAVICA L, I, 2, 8, II, 1 3, FP ST A 

16, 17 

'Asclepias sericea' 2 FP 

I cannot trace this name. A. serica Raf., a synonym 

of A. syriaca L., may be intended. 
♦Calotropis procera (Aiton) Aiton f. 2, 17 FP 

Cynanchum adalinae (K. Schum.) K. Schum. subsp. 

ADALINAE 2, 1 7 FP 

GONGRONEMA LATIFOLIUM Benth. ST 

Marsdenia latifolia (Benth.) K. Schum. i, 25, 26 

Marsdenia exellii C. Norman i, 14 ST 

Oncostemma cuspidatum K. Schum. i FP ST 

FP : Bokoko, Mildbraed 6g48. Not recorded from FP 

in 17. 
Rhynchostigma racemosum Benth. 17 FP 

Tylophora oblonga N. E. Brown 2 FP 

Not recorded from FP in 17. The type {Mann 2yy) is 

from FP. 
Tylophora oculata N. E. Brown 2 FP 

Not recorded from FP in 17. 
Tylophora sylvatica Decaisne 2, 17 FP 

107. RUBIACEAE 

AiDiA GENiPiFLORA (DC.) Dandy 

Randia genipi flora DC. 
AiDiA FALLENS (Hiern) G. Taylor 
AiDiA QuiNTASii (K. Schum.) G. Taylor 
AiDiA WATTii G. Taylor 
Anthospermum asperuloides Hook. f. 
Argostemma africanum K. Schum. 

FP : Mildbraed 6308. Not recorded from FP in 17. 
AuLACOCALYX LE-TESTUi (Pellcgrin) E. Petit 

Dorothea le-testui Pellegrin 
Belonophora coffeoides Hook. f. 
Bertiera annobonensis G. Taylor 
Bertiera laxa Benth. var. laxa 
Bertiera pedicellata (Hiern) Wernham 
Bertiera racemosa (G. Don) K. Schum. var. racemosa 

B. montana Hiern 
Bertiera retrofracta K. Schum. 

B. obversa K. Krause 
BoRRERiA OCYMOIDES (Burm. f.) DC. 

BORRERIA PRINCEAE var. PUBESCENS HeppCt 

BoRRERiA scABRA (Schumach.) K. Schum. 

B. monticola Mildbr. 
BoRRERiA VERTiciLLATA (L.) G. F. W. Meyer 
Canthium dunlapii Hutch. & Dalziel 
Canthium henriquesianum (K. Schum.) G. Taylor 
Canthium setosum Hiern 
Canthium subcordatum DC. 

C. glabriflorum Hiern 
Canthium sp. aff. anomocarpum DC. 



17 






FP 






2 












I 








ST 


E 


I 








ST 


E 


I 








ST 


E 


2, 17 






FP 
FP 


ST 




I 












I 








ST 


E 


I, 16 








A 


E 


2, 17 






FP 






I 








ST 


E 


I. 13. 17 






FPPST 




2 












2, 17 






FP 






2 












I, 8, 17 






FP P ST A 




17 






FP 






17 






FP 






2 












I, 8, II, 


13. 


14 




PST 




17 






FP 






I. 17 








ST 




17 






FP 


ST 




I, 2, II, 


13. 


25 








I 








P 





OF THE GULF OF GUINEA 



373 



Canthium sp. aff. glabriflorum Hiern 

Canthium sp. 

Canthium sp. 

Cephaelis mannii (Hook, f.) Hiern 

Uragoga mannii (Hook, f.) Hutch. & Dalziel 
Cephaelis peduncularis Salisb. 

Uragoga membranifolia Mildbr., nom. nud. 
Cephaelis peduncularis var. hypsophila (Schum. & 

Krause) Hepper 
Cephaelis spathacea Hiern 
Chassalia cristata (Hiern) Bremek. 
Chassalia doniana (Benth.) G. Taylor 
Chassalia hiernii (Kuntze) G. Taylor var. hiernii 
Chassalia hiernii var. glandulosa G. Taylor 
Chassalia kolly (Schumach.) Hepper 

Psychotria benthamiana Hiern 
Chassalia subherbacea (Hiern) Hepper 

Psychotria subherbacea Hiern 
♦Cinchona 

Numerous cultivated species are grown in ST ; see i. 
♦Coffea arabica L. 

♦Coffea liberica Bull ex Hiern 
Coffea macrochlamys K. Schum. 
♦Coffea stenophylla G. Don 

Many other introduced species and varieties of Coffea 

are listed in 2 and 13. 
Corynanthe sp. ? 

FP : Bokoko, Burchardt. 
Craterispermum montanum Hiern 
Cremaspora thomsonii Hiern 

FP : Burchardt. Not recorded from FP in 17. 
Cuviera subuliflora Benth. 

DiCTYANDRA ARBORESCENS Welw. ex Hook. f. 

DioDiA SCANDENS Swartz 
DiODiA SERRULATA (Beauv.) G. Taylor 
EcPOMA CAULiFLORUM (Hiern) N. Halle in Adansonia 
3: 173 (1963). 

Sabicea cauliflora Hiern 
ST : Rose 208, fide N. Halle (additional record). 
EcPOMA GiGANTisTiPULUM (K. Schum.) N. Halle, loc. cit. 
Sabicea gigantistipula K. Schum. 
P : Rose 419. fide N. Hall6. 
Galium simense Fresen. 
Galium thunbergianum Ecklon & Zeyher 

G. biafrae Hiern 
Gardenia nigrificans Wernham 

Not in 17. 
Geophila afzelii Hiern 

G. cordiformis A. Chev. ex Hutch. & Dalziel 
Carinta cordiformis (A. Chev. ex Hutch. & Dalziel) 
G. Taylor 
Geophila neurodictyon (K. Schum.) Hepper 
Psychotria neurodictyon K. Schum. 



13 








ST 


I 








P 


13 








ST 


17 








FP 


2 










I, 13 








PST A 


2 










17 








FP 


I, II 


13 






ST 


17 








FP 


I, II 








ST 


I, II 








P 


I 








P 


17 








FP 


2 










17 








FP 


2 










I, 2, 


II, 


13, 


17. 


FP PST 


25 










I, 2, 


13, 


16, 


25 


FP P ST A 


17 








FP 


I, 2, 


II. 


13. 


25 


ST 



FP 



1, 13 


PST A 
FP 


E 


2, 17 


FP 




7. 17 


FP 




17 


FP 




I 


PST 






ST 


E 



I, II 

17 
17 

17 
17 

2 

2 

17 

I 



FPP 



FP 
FP 

FP 



II, 
16 
I, 2 



14 



374 



ANGIOSPERMS OF THE ISLANDS 



GfiOPHiLA OBVALLATA (Schumacher) Didrichsen 

Carinta ohvallata (Schumacher) G. Taylor 
Geophila repens (L.) I. M. Johnston 

G. uniflora Hiern 

Carinta uniflora (Hiern) G. Taylor 
Heinsia crinita (Afz.) G. Taylor var. crinita 

H. pulchella K. Schum. 
? Heinsia epidendron Mildbr. 

Not in 17. Probably an error for Hymenodictyon 
epidendron Mildbr. (= H. biafranum Hiern). 
Hymenodictyon biafranum Hiern 
Ixora breviflora Hiern 

*IXORA COCCINEA L. 

Ixora foliosa Hiern 

FP : Moka, Tessmann 28g6. Not recorded from FP 

in 17. 
Ixora guineensis Benth. 

/. riparia Hiern 
Ixora nematopoda K. Schum ' 

I. rosea K. Schum. 
Ixora thomeana (K. Schum.) G. Taylor 
Lasianthus africanus Hiern 
Lasianthus aff. mannii Wernham 
Massularia acuminata (G. Don) Bullock ex Hoyle 

Randia acuminata (G. Don) Benth. 
MiTRAGYNA ciLiATA Aubr^v. & Pellegrin 
Mitriostigma barteri Hook. f. ex Hiern 
MoRiNDA longiflora G. Don 
MoRiNDA LUCiDA Benth. 
MoRiNDA MORiNDOiDES (Bak.) Milne-Rcdh. 

M. confusa Hutch. 
MussAENDA ARCUATA Lam. ex Poiret 
MussAENDA ELEGANS Schumacher 

Not recorded from FP in 17. 
MussAENDA ERYTHROPHYLLA Schumacher 

Not recorded from FP in 17. 

MussAENDA ISERTIANA DC. 

MussAENDA LANCiFOLiA K. Krause 
MUSSAENDA POLiTA Hiem 

FP : Bokoko, Mildbraed 6814. Not recorded from 

FP in 17. 

MuSSAENDA TENUIFLORA Benth. var. TENUIFLORA 

MussAENDA TENUIFLORA var. PRiNCiPENSis G. Taylor 

MussAENDA TENUIFLORA var. THOMENSIS G. TayloT 

MUSSAENDA aff. AFZELii G. Don 
Nauclea latifolia Smith 

Sarcocephalus esculentus Afzel. ex Sabine 
Not recorded from FP in 17. 
Oldenlandia corymbosa L. 
Oldenlandia herbacea (L.) Roxb. 
Oldenlandia lancifolia (Schumacher) DC. 
Otomeria cameronica (Bremek.) Hepper 

O. guineensis sensu G. Taylor 

Tapinopentas latifolia Verde. 



I, 2, 17 


FPP 


II 






PST 


I 




II 




I. 17 


FPP 


2 




2 


FP 


I. 17 


FPP 


2, 17 


FP 


II, 16 


ST 



FP 



17 


FP 




2 






17 


FP 




2 






I 


ST 


E 


I. 15 


PST 


E 


17 


FP 




17 


FP 




2 






17 


FP 




2, 17 


FP 


E 


2, 17 


FP 




I, 2, 17, 25, 26 


FPP ST 




17 


FP 




2 






2, 17 


FP 




2 


FP 





FP 



2, 17 


FP 


2, 17 


FP 




FP 


2, 17 


FP 


I 




I 




13 





ST 



E 



FP 



I, 2, II, 17 


FP P ST A 


16 


A 


17 


FP 


17 


FPP 


1 




IX 





OF THE GULF OF GUINEA 



375 



OxYANTHUs FORMOsus Hook. f. ex Planchoii 

O. hreviflorus Benth. 
OxYANTHUs GRACILIS Hiem 

OxYANTHUS SPECIOSUS DC. 
OXYANTHUS Sp. A 

Parapentas setigera (Hiern) Verde. 

Virecta setigera Hiern 
Pauridiantha camposii (G. Taylor) Exell, comb. nov. 
Urophyllum camposii G. Taylor in Exell, Cat. Vase. 
PI. S. Tomi : 202 (1944). 
Holotype : Campos yi (COI). 
Pauridiantha canthiiflora Hook. f. 

Urophyllum, canthiiflorum (Hook, f.) Hutch. & 
Dalziel 

Pauridiantha floribunda (Schum. & Krause) Bremek. 
Urophyllum floribundum Schum. & Krause 
U. annohonense Mildbr., nom. nud. 
Pauridiantha hirtella (Benth.) Bremek. 

FP : Bokoko, Mildbraed 6g38. Not recorded from 
FP in 17. 
Pauridiantha insularis (Hiem) Bremek. 

Urophyllum insulare Hiern 
Pauridiantha paucinervis (Hiern) Bremek. 

Urophyllum paucinerve Hiern 
Pauridiantha rubens (Benth.) Bremek. 

Urophyllum rubens Benth. 
Pavetta bidentata Hiern 
Pavetta brachycalyx Hiern 

FP : Bokoko, Mildbraed 6812. Not recorded from 
FP in 17. 

Pavetta corymbosa (DC.) F. N. WiUiams 
P. nitida (Schumach.) Hutch. & Dalziel 
Recorded from FP in Fl. W. Trop. Afr., ed. i, but not 
in 17. 

Pavetta glaucescens Hiern 
Pavetta hispida Hiern 
Pavetta hookerana Hiem 

P. exellii Bremek. 
Pavetta monticola Hiem 

P. dermatophylla Mildbr., nom. nud. 
Pavetta neurocarpa Benth. 
Pavetta oresitropha Bremek. 
Pavetta owariensis Beauv. 
Pavetta rigida Hiem 

Pentas schimperana subsp. occidentalis (Hook, f.) 
Verde. 

P. occidentalis (Hook, f.) Benth. & Hook, ex Hiern 
Pentodon pentandrus (Schumacher) Vatke 

Oldenlandia macrophylla DC. 
Polycoryne fernandensis (Hiern) Keay 
Gardenia fernandensis Hiern 

POUCHETIA CONFERTIFLORA Mildbr. 
POUCHETIA PARVIFLORA Benth. 



2, 17 


FP 


2 




2, 17 


FP 


I, 9, II, 13, 17 




17 


FP 


17 


FP 


2 





17 

2 

17 
I, '■ 
2 



13 



2, 17 




2, 17 




17 




2 




I, II, 


13 


2 




2, 17 




2, 17 




17 




2, 17 




17 




I, II, 


13 


17 




I 




17 




2 




I, 2 




I, 2, ] 


C7 



ST A 



ST 



FP 



PST A 



FP 







ST 


1. 13 

17 


FP 




2 
17 


FP 


ST 


1. 2 

2, 17 

2 


FP 
FP 





FP 



FP 
FP 
FP 



ST A 



FP 




FP 




FP 




FP 




FP 


ST 




ST 


FP 




FPPST 



376 



ANGIOSPERMS OF THE ISLANDS 



worn, nud., non 
epiphytica Elmer 



Not in 17. 



2, 17 



1, 13 




ST 


17 


FP 


ST? 


I 






2 






I, 13 




ST 



Preussiodora sulphurea (K. Schum.) Keay 17 

Randia exserta K. Schum. 2 

? PsiLANTHUS BiCARiNATUS Mildbf., nom. nud. 2 

Perhaps an error for Psychotria hicarinata Mildbr. 

( = Psychotria recurva Hiern) . 
? PsiLANTHUS EPiPHYTicuM Mildbr., nom. nud. 2 

Probably an error for Psychotria epiphytica Mildbr. 

PsiLANTHUS MANNII Hook. f. 

Heinsia jasminiflora sensu A. Chev. 
Psychotria bifaria Hiern 
Psychotria brachyantha Hiern 
Psychotria calva Hiern 
Psychotria cornuta Hiern 
Psychotria crassicalyx K. Krause 
Psychotria epiphytica Mildbr., 

P. epiphytica K. Krause nee P 

FP : Mildbr aed 6433, 7i8g. 
Psychotria erythropus K. Schum. 

FP : Mildbr aed 6327. Not recorded from FP in 17. 
Psychotria euchlora K. Schum. 
Psychotria fernandopoensis E. Petit 
? Grumilea sp. aff. sphaerocarpa Hiern 
Psychotria sphaerocarpa (Hiern) Hutch. & Dalziel, 
non P. sphaerocarpa Wall. 
Psychotria guerkeana K. Schum. 

Psychotria henriquesiana K. Schum. i 

Psychotria hierniana Exell, nom. nov. 

Grumilea macrocarpa Hiern in Oliver, Fl. Trop. Afr. 
3 : 217 {1877), non Psychotria macrocarpa Hook. f. 
(1880). I 

Holotype : Mann (K), from S. Tome 
? Psychotria insidens Hiern 2, 17 

E. Petit considers this to be a new genus of Rubiaceae. 

The specimen of this species recorded in 17 as Mann 

301 has been renumbered Mann 301a, as there is also 

a specimen of P. pteropetala numbered Mann 310. 
Psychotria latistipula Benth. 
Psychotria leptophylla Hiern 

P. setacea Hiern 
Psychotria lucens Hiern 
Psychotria malchairei De Wild. 
Psychotria molleri K. Schum. 
Psychotria mollipes K. Krause 

FP : above Basils. Mildbraed 6360 
Psychotria nubicola G. Taylor 
Psychotria principensis G. Taylor 
Psychotria pteropetala K. Schum. 
Psychotria recurva Hiern 

Psilanthus carinatus Mildbr., nom. nud. 
Psychotria sciadephora Hiern 

FP : Bokoko, Mildbraed 6853. Not recorded from 

FP in 17. 
Psychotria subobliqua Hiern i, 2, 11, 13, 17 

Psychotria thomensis G. Taylor i 



FP 
FP 

FP 



17 


FP 




2 






2, 17 


FP 




2, 17 


FP 




2, 17 


FP 




2, 17 


FP 


E 


2, 17 


FP 


E 



FP 



FP 



FP 



ST 
ST 



2, 17 


FP 




17 


FP 




2 






I 


P 


E 


17 


FP 




I 


ST 


E 


2 


FP 




I, 13 


ST 


E 


I 


P 


E 


2, 17 


FP 




2, 17 


FP 





FP 



FP P ST 
ST 



E 



OF THE GULF OF GUINEA 



377 



PsYCHOTRiA VENOSA (Hiem) E. Petit 
Grumilea venosa Hiern 

PSYCHOTRIA Sp. 

FP : Musola, Mildbraed 7042, as Grumilea sp. 

ROTHMANNIA LONGIFLORA Salisb. 

Randia maculata DC. 
RoTHMANNiA MACROCARPA (Hiem) Keay 
Randia myrmecophylla De Wild. 
Not in 17. 
RoTHMANNiA OCTOMERA (Hook.) Fagerl. 

Randia octomera (Hook.) Hook. f. 
ROTHMANNIA URCELLiFORMis (Schwcinf . ex Hiem) Bullock 
ex Robyns 

R. arcuata Bremek. 
? ROTHMANNIA WHiTFiELDii (Lindl.) Dandy 

FP : Bokoko, Mildbraed 6g4y, identification doubtful. 
Not recorded from FP in 17, 

RUTIDEA DECORTICATA Hiem 

FP : Bokoko, Mildbraed 6822 ; Musola, Mildbraed 

7012. Not recorded from FP in 17, but perhaps these 

specimens are R. syringoides. 
RuTiDEA SMiTHii Hiem 

The Annobon plant may be R. smithii var. welwitschii. 
RuTiDEA SMITHII var. WELWITSCHII Scott Elliot 
RuTiDEA SYRINGOIDES (Webb) Bremek. 
Rytigynia concolor (Hiern) Robyns 
Sabicea africana (Beauv.) Hepper 
Stipularia africana Beauv. 

Not recorded from FP in 17. 
Sabicea calycina Benth. 
Sabicea capitellata Benth. 
Sabicea exellii G. Taylor 
Sabicea hierniana Wernham 

FP : Mildbraed 6288. Not in 17. 
Sabicea ingrata K. Schum. 
Sabicea insularis (Wernham) G. Taylor 

5. annobonensis Mildbr., nom. nud. 
Sabicea schaeferi Wernham 

Also in Cameroon fide N. Halle. 
Sabicea urbaniana Wernham 
Sabicea urceolata Hepper 
Sabicea aff. cauliflora Hiern 

Presumably a species of Ecpoma. 
Sacosperma paniculatum (Benth.) G. Taylor 

Oldenlandia peltospermum Hiern 
ScHUMANNioPHYTON magnificum (K. Schum.) Harms 

FP : Bokoko, Mildbraed 6350, 6g36. Not recorded 

from FP in 17. 
Tarenna bipindensis (K. Schum.) Bremek. 

Ixora bipindensis (K. Schum.) K. Schum. ex Hutch. 
& Dalziel 
Tarenna conferta (Benth.) Hiern 
Tarenna grandiflora (Benth.) Hiern 
Tarenna nitiduloides G. Taylor 



17 

I. 2, 



17 

2 



17 

2 

15 
II 



II, 25 



I. 17 

15 
17 

I 



17 

2 

2, 17 
2, 17 
I. 15 



FP ST 
FP 
FP 
FP 

FP 



ST 



FP? 



FP 



ST 



FP 



FP 



A? 



2, 17 


FP 






2, 17 


FP 






I 


FP 


ST 


E 


I 




ST 


E 


£> 




ST A 


E 


2 








17 


FP 




E? 


2, 17 


FP 




E 


17 


FP 






13 


P 






I, II, 17 


FPPST 




2 


FP 







FP 



FP 
FP 



ST 



378 



ANGIOSPERMS OF THE ISLANDS 



Tarenna sp. I I 

Tarenna sp. 2 I 

Tarenna sp. 3 I 

Tricalysia biafrana Hiern 

FP : Bokoko, Mildbraed 6g62. Not recorded from 

FP in 17. 
Tricalysia oligoneura K. Schum. 

FP : Bokoko, Mildbraed 6gog ; a doubtful record. 

Not recorded from FP in 17. 
Tricalysia fallens Hiern 2, 17 

Uncaria AFRICAN a G. Don var. African a 16 

? Urogoga nutans K. Krause 2 

Presumably either Cephaelis mannii or C. peduncularis. 
ViRECTARiA MAJOR (K. Schum.) Verdc. var. major 17 

ViRECTARiA PROCUMBENS (Smith) Bremek. 17 



FP 

FP? 

FP 

FP 

FP 
FP 



ST 
ST 



108. caprifoliaceae 



♦Sambucus canadensis L. 

FP : Santa Isabel, Mildbraed 6483. 



FP 



109. COMPOSITAE 

♦Acanthospermum hispidum DC. 
Adenostemma mauritianum DC. 
Adenostemma perrottetii DC. 

Ageratum conyzoides L. 

Ambrosia maritima L. 

AspiLiA RUDis Oliver & Hiern subsp. rudis 

Bidens pilosa L. 

Blumea aurita (L. f.) DC. 

Not recorded from FP in 17. 
♦Chrysanthemum indicum L. 
♦Cichorium intybus L. 
Conyza bonariensis (L.) Cronq. 

Erigeron sumatrensis Retz. 
CoNYZA CLARENCEANA (Hook. f .) Oliver & Hiern 
Conyza floribunda Kunth 

Erigeron floribundus (Kunth) Schultz Bip. 
CoNYZA persicifolia (Benth.) Oliver & Hiern 

Eschenbachia persicifolia (Benth.) Exell 
Conyza steudelii Schultz Bip. ex A. Rich. 

FP : Escarri k-sy. Not recorded from FP in 17. 
CoNYZA theodori R. E. Frics 

FP : Escarrd e-2. Not recorded from FP in 17. 
Crassocephalum biafrae (Oliver & Hiern) S. Moore 

Senecio biafrae Oliver & Hiern 
Crassocephalum bougheyanum C. D. Adams 
Crassocephalum crepidioides (Benth.) S. Moore 

Gynura crepidioides Benth. 
Crassocephalum gracile (Hook, f.) Milne-Redh. ex 
Guinea 



1, 25 P ST 

2, 17 FP 

I, 6, II, 13, 16, FP ST A 

17. 25 
I, 2, 6, 9, II, 13, FPPST A 

17, 25, 26 
I PST 

1 P 

I, 6, 8, 9, II, 13, FPPST A 
17, 25, 26 

2 FP 



I 






ST 


I 






ST 
PST 


I 








3. 6, 17 






FP 
FP 


6, 17 








2, 6, 17 






FP ST 


I, 8, II. 


13. 


25 


FP 
FP 


6, 17 






FP ST 


I 








17 






FP 


6, II, 15, 


. 17 




FP 


I 









6, 17 



FP 



OF THE GULF OF GUINEA 



379 



Crassocephalum mannii (Hook, f.) Milne-Redh. 

Senecio mannii Hook. f. 
Crassocephalum montuosum (S. Moore) Milne-Redh. 
Crassocephalum picridifolium (DC.) S. Moore 
Crassocephalum rubens (Juss. ex Jacq.) S. Moore 

FP : Moka, Tessmann 28^2. Not recorded from FP 

in 17. 
Crassocephalum vitellinum (Benth.) S. Moore 

Gynura vitellina Benth. 
Dichrocephala chrysanthemifolia (Blume) DC. 
Dichrocephala integrifolia (L. f.) Kuntze 

D. bicolor (Roth) Schlechtendal 
Eclipta prostrata (L.) L. 
Elephantopus mollis Kunth 

♦Eleutheranthera ruderalis (Swartz) Schultz Bip. 
Emilia coccinea (Sims) G. Don 

E. flammea Cass. 
E. sagittata DC. 

Emilia praetermissa Milne-Redh. 

FP : fide Escarr6. Not recorded from FP in 17. 
Emilia sonchifolia (L.) DC. 
Enydra fluctuans Lour. 
♦Epaltes brasiliensis DC. 
Ethulia conyzoides L.f. 

FP : Melville 6- Wrigley 62y (K) . Not recorded from 

FP in 17. 
♦Galinsoga ciliata (Raf.) Blake 

FP : fide Escarre. 
•Galinsoga parviflora Cav. 
Gnaphalium luteo-album L. 
Helichrysum cymosum (L.) Less. 

H. fruticosum Vatke 
Helichrysum foetidum (L.) Moench 
Helichrysum globosum Schultz Bip. ex A. Rich. 
Helichrysum mannii Hook. f. 
Helichrysum odoratissimum (L.) Less. 
Achyrocline hochstetteri Schultz Bip. 
Lactuca capensis Thunb. 

Lactuca glandulifera Hook. f. var. glandulifera 
Laggera alata var. Montana C. D. Adams 
Melanthera scandens (Schumacher) Roberty 

M. brownei (DC.) Schultz Bip. 
Microglossa densiflora Hook. f. 
Microglossa pyrifolia (Lam.) Kuntze 
MiKANiA coRDATA (Burm. f.) B. L. Robinson var. 

CORDATA 

M. scandens sensu auctt. 
Mikania sp. 

M. scandens sensu Exell 
MiKANiopsis paniculata Milne-Redh. 

FP : Melville S- Wrigley 625 (K). Not recorded 
from FP in 17. 
Senecio clarenceanus Hook, f 



6, 17 


FP 




2 






15- 17 


FP 


ST 


6, 17 


FP 
FP 




6, 17 


FP 




2 






2, 6, 17 


FP 




16, 17 


FP 


ST 


I, II, 13 






I 




PST A 


I, 6, II, 13, 17, 


FPPST 


25 






16, 17 


FP 


A 


17 


FP 


ST 


13 






2, 13 


FP 




I, 2, 6, 14, 17 


FPPST 


2, 17 


FP 




I 




PST 


I, 17 


FP 


ST 



FP 

I ST 

1, 6, 13, 14, 17 FP P ST 
17 FP 

2, 6 

1, 2, 6, 17 FP ST 

2, 6, 17 FP 
2, 6, 17 FP 
17 FP 
2, 6 

6, 17 FP 

6, 17 FP 

6, 17 FP 

17 FP 

2, 6 
6, 17 

I, 15. 17 

II, 16, 17 

I, 2, 6, 7 
II 
I 

II. 15. 17 



2, 6, 17 



FP 

FPPST 

FPP A 
ST A 
FP ST 

FP 



38o 



ANGIOSPERMS OF THE ISLANDS 



SONCHUS OLERACEUS L. 

SoNCHUs scHWEiNFURTHii Oliver & Hiern 

FP : fide Escarr6. Not recorded from FP in 17. 
Spilanthes filicaulis (Schumacher) C, D. Adams 

5. acmella sensu Exell 

S. caulirhiza sensu Exell 
Struchium sparganophora (L.) Kuntze 

Sparganophorus vaillantii Crantz 
Synedrella nodiflora (L.) Gaertn. 

♦Tagetes erecta L. 

Triplotaxis stellulifera (Benth.) Hutch. 

Vernonia amygdalina Delile 

? V. senegalensis sensu Guinea 
Vernonia biafrae Oliver & Hiern 
Vernonia blumeoides Hook. f. 

Not recorded from FP in 17. 
Vernonia conferta Benth. 
Vernonia insignis (Hook, f.) Ohver & Hiern 

FP : Escarrd e-2g. Not recorded from FP in 17. 
Vernonia mokaensis Mildbr. & Mattf . 
Vernonia myriantha Hook. f. 
♦Zinnia peruviana (L.) L. 

no. gentianaceae 
Canscora decussata (Roxb.) J. A. & J. H. Schultes 
ExACUM quinquenervium Griseb. 
Neurotheca loeselioides (Spruce ex Progel) Baillon 
Sebaea brachyphylla Griseb. 

5. multinodis N. E. Brown 
Swertia abyssinica Hochst. 

S. clarenceana Hook. f. 
'Swertia camerunensis' 
I cannot trace this name. 



I, 6, 13, 17 




FP 


ST 


FP 








17 




FPPST 


I, 6 








II 








I, 6, 9, II, 


13. 


FPPST 


17, 25, 26 








2. 7 








I, 6, II, 13, 


16, 


FP P ST A 


17, 25 








6 




FP 




I, 2, 6, II, 


13, 


FPPST 


17 








I, 2, 6, 13, 


17, 


FP 


ST A 


25, 26 








2 








2, 6, 17 




FP 




6 




FP 




2, 6, 7, 17 




FP 
FP 




2, 6, 17 




FP 




2, 17 




FP 




1.6 




FP 


ST 


16 






A 


16 






A 


2. 17 




FP 




17 




FP 




2. 7 








17 




FP 




2.7 








7 




FP 





III. PRIMULACEAE 

Ardisiandra sibthorpioides Hook. f. 2, 7, 17 



♦Plumbago auriculata Lam. 

P. capensis Thunb. 
Plumbago zeylanica L. 



112. plumb agin ACE AE 



I, II 

1. 17 



FP 

ST 

FP ST 



♦Plant ago major L. 
Plantago palmata Hook, f . 



113. plantaginaceae 



1. 15 

2, 17 



P? ST 



FP 



114. campanulaceae 

Cephalostigma perrottetii a. DC. 16 

Wahlenbergia arguta Hook. f. 2, 17 

Wahlenbergia mannii Vatke 2, 17 



FP 
FP 



17 


FP 






2, ly 


FP 




E 


I, 13, 15 




ST 


E 


2, 7. 17 


FP 






17 


FP 






I. 17 


FP 


ST 





OF THE GULF OF GUINEA 381 

115. LOBELIACEAE 

DiELSANTHA GALEOPSOiDES (Engl. & Diels) E. Wimmer 

Lobelia acutidens Hook. f. 

Lobelia barnsii Exell 

Lobelia columnaris Hook. f. 

Lobelia kamerunensis Engl, ex Hutch. & Dalziel 

Lobelia molleri Henriq. 

MoNOPSis STELLARioiDES var. scHiMPERANA (Urban) E. 

Wimmer 17 FP 

116. GOODENIACEAE 
SCAEVOLA PLUMIERI (L.) Vahl I ST 

117. boraginaceae 
CoRDiA aurantiaca Baker 2, 17 FP 
♦CoRDiA gerascanthus L. 13, 27 ST 
CoRDiA PLATYTHYRSA Baker 27 FP ST 

C. odorata Giirke 

Cordia sp. 25 

FP : Bokoko, Mildbraed 6g3g. Not recorded from 

FP in 17. 
CORDIA senegalensis Juss. FP 

FP : San Carlos, Mildbraed 6g32. Not recorded from 

FP in 17. 
Cynoglossum lanceolatum Forssk. 2 

Not recorded from FP in 17. 
Ehretia scrobiculata Hiern i 

Ehretia zenkeri Giirke ex Baker & Wright 11, 27 

E. scrobiculata sensu Espirito Santo 25 

Treated in 17 as a variety of E. cymosa Thonn. 
Heliotropium indicum L. I, II, 13, 25 

FP : fide Escarre. Not recorded from FP in 17. 
Myosotis abyssinica Boiss. & Renter 2, 17 

118. SOLAN aceae 
♦BrOWALLIA AMERICANA L. II 

B. demissa L. 8 

♦Brunfelsia uniflora (Pohl) D. Don 

Franciscea uniflora Pohl 13 

Capsicum annuum L. i, 2, 16, 25 

Not recorded from FP in 17. 

Capsicum frutescens L. i, 2, 13, 25 

Not recorded from FP in 17. 

♦Cestrum laevigatum var. puberulum Sendtner i, 11, 13, 25 

Now quite spontaneous in ST. 

♦Cyphomandra betacea (Cav.) Sendtner i 

♦Datura CANDIDA (Pers.) Saff. 11, 16 

FP : Moka, Exell jy'j. Common as a bush in the grass- 
land and as a small tree in the woodland. Used by the 
Bubis to make fences for their villages. 

♦Datura metel L. i, 2, 25, 26 FP ST 

♦Datura stramonium L. 2 FP 

Discopodium penninervium Hochst. 2, 17 FP 



FP 






P E 




ST 


FPPST 


FP 






ST 




ST 


FP 


ST A 


FPPST 


FPPST 




ST 


FP 


ST A 



382 



ANGIOSPERMS OF THE ISLANDS 



♦Lycopersicum esculentum var. cerasiforme (Dunal) 
Alef. 

L. cerasiforme L. 

Solatium lycopersicum L. 
♦NiCANDRA PHYSALOiDES (L.) Gaertn. 

Physalis peruviana sensu Guinea 
*nlcotiana tabacum l. 
♦Physalis angulata L. 
Physalis micrantha Link 

SOLANUM ACULEATISSIMUM Jacq. 

SoLANUM DASYPHYLLUM Schumachcr 
S. duplosinuatum Klotzsch 
Not recorded from FP in 17. 

SoLANUM INDICUM L. 

SoLANUM INDICUM var. grandemunitum Bitter 
SoLANUM INDICUM var. MODiCEARMATUM Bitter 

SOLANUM MACROCARPON L. 

Not recorded from FP in 17. 

♦SOLANUM MELONGENA L. 

SoLANUM NIGRUM L. scns. lat. 
S. nodiflorum Jacq. 
The specimens from our area are probably all referable 
to 5. americanum Mill. 

SOLANUM TERMINALE Subsp. INCONSTANS (C. H. Wright) 

Heine 

5. inconstans C. H. Wright 

SOLANUM TERMINALE ForSSk. Subsp. TERMINALE 

S. bifurcatum Hochst. ex A. Rich. 

SOLANUM TERMINALE Subsp. WELWITSCHII (C, H. 

Wright) Heine 

5. welwitschii var. strictum C. H. Wright 
SoLANUM TORVUM Swartz 
♦solanum tuberosum l. 

119. convolvulaceae 
♦Ipomoea alba L. 

Calonyction aculeatum (L.) H. D. House 
C. hona-nox (L.) Bojer 
Not recorded from FP in 17. 
Ipomoea batatas (L.) Lam. 

Not recorded from FP in 17. 
Ipomoea cairica (L.) Sweet 
Ipomoea eriocarpa R. Br. 

/. hispida (Vahl) Roemer & Schultes, non Zuccagni 
Not recorded from FP in 17. 
♦Ipomoea hederifolia L. 

Quamoclit coccinea sensu Exell 
Ipomoea involucrata Beauv. 
Ipomoea mauritiana Jacq. 
/. digitata sensu Exell 
? Convolvulus paniculatus sensu Guinea 
Ipomoea nil (L.) Roth 

/. hederacea sensu Baker & Rendle 
FP : Buchholz. Not recorded from FP in 17. 



I, 2, 


16 




FP P ST A 


2 








2 








I, II 


. 25. 


26 


FPPST 


2 








I, 2, 


16 




FP ST A 


I, 2, 


13. ■ 


[6, 17 


FP P ST A 


16 






A 


I 






P 


16 






FPP A 


I, 2 








I. 2 






FP ST . 


17 






FP 


17 






FP 


I. 2 






FP ST 


I. 2 






FP ST 


1, 2, 


8,9. 


II, 17 


FP P ST A 


16 









17 

2. 7 



FP 



ST 



17 


FP 


2 

'2, 17 

I 


FP 

ST 


16 


FP P ST A 


I 





I, 2, II, 13 



9, 



17 

II 



1, II 

2, 16, 17 
15. 16, 17 
I, 2, 7 

2 
16 



FP P ST 

FPPST 
FP ST A 



FPPST 

FP A 

FP ST A 



FP 



OF THE GULF OF GUINEA 383 

Ipomoea pes-caprae subsp. BRASiLiENSis (L.) Van 1,2,7,9,11,13, FP P ST A 

Ooststr. 1 7 

♦Ipomoea purpurea Roth 2 FP 

Ipomoea quamoclit L. 2, 17 FP P ST 

Quamoclit vulgaris Choisy i, 11 

Ipomoea rubens Choisy FP ? ST 

I. lilacina Blume, non Schrank 2 

/. riparia G. Don i 

Not recorded from FP in 17, but the FP record is 

doubtful. 

♦Ipomoea setifera Poiret i P 

Ipomoea stolonifera (Cyr.) J. F. Gmel. i, 7 FP ST 

Not recorded from FP in 17. 

♦Ipomoea tiliacea (Willd.) Choisy i, 17 P 

♦Ipomoea tuberosa L. 2 FP 

Merremia aegyptia (L.) Urban i ST 

♦Merremia dissecta (Jacq.) HaUier f. 2, 17 FP 

Merremia pterygocaulos (Choisy) HaUier f. 2 FP 

FP : Buchholz. Not recorded from FP in 17. 
Merremia tridentata subsp. angustifolia (Jacq.) 

Van Ooststr. 11 ST 

M. angustifolia (Jacq.) Halher f. i 

Merremia umbellata (L.) HaUier f. subsp. umbellata i, 2, 11, 17 FP P 

120. SCROPHULARIACEAE 

Alectra sessiliflora var. monticola (Engl.) Melch. 16, 17 FP P A 

A. communis Hemsley i 

Celsia densifolia Hook, 17 FP 

' Celsia densiflora Hook, f.' 2 

♦Lendneria verticillata (Mill.) Britton FP 

Stemodia verticillata (Mill.) Boldingh 17 

LiMOSELLA australis R.Br. FP 

L. subulata Ives 

L. tenuifolia Nutt. 
LiNDERNiA DIFFUSA (L.) Wettst. var. diffusa 
LiNDERNiA DIFFUSA var. PEDUNCULATA (Benth.) Skan 

L. stictantha (Hiern) Skan 
LiNDERNiA vogelii Skan 
♦Russelia equisetiformis Schlectendal & Cham. 

R. juncea Zucc. 

SCOPARIA DULCIS L. 
SiBTHORPIA EUROPAEA L. 

S. australis Hutch. 
Thunbergianthus quintasii Engl. 

TORENIA DINKLAGEI Engl. 

T. parviflora sensu Guinea 
ToRENiA THOUARSii (Cham. & Schlecht.) Kuntze 
Veronica mannii Hook. f. 



121. lentibulariaceae 



Utricularia mannii Oliver 

U. bryophila Ridley 
Utricularia striatula Smith 



17 






2 






I, 16, 


17 


FP P ST A 


I, II, 


16, 17 


ST A 


8 






17 




FP 


II, 13 




PST 


8 






I. 25 




PST A 


17 




FP 


2 






I 




ST 


17 




FP 


2 






I 




PST 


2, 17 




FP 


14. 15 


17 


FPPST 


I. 13 






I, 15. 


17 


FPP 



384 



ANGIOSPERMS OF THE ISLANDS 



122. GESNERIACEAE 

ACANTHONEMA STRIGOSUM Hook. f. 
DiDYMOCARPUS KAMERUNENSIS Engl. 

Epithema tenue C. B. Clarke 
Streptocarpus elongatus Engl. 

FP : fide Escarr6. Not recorded from FP in 17. 
Streptocarpus insularis Hutch. & Dalziel 

S. denticulatus Engl., non Turrill 
Streptocarpus nobilis C. B. Clarke 
Streptocarpus thomensis Exell 



2, 17 

17 

I, 2, 17 

7. 17 
2 

14. 17 
I. 13 



FP 

FP 

FP ST 

FP 



FP 



E 



ST 

ST 



123. bignoniaceae 

♦Jacaranda mimosifolia D. Don 
KiGELiA africana (Lam.) Benth. 

K. impressa Sprague 

K. pinnata (Jacq.) DC. 
Markhamia lutea (Benth.) K. Schum. 
Newbouldia laevis (Beauv.) Seem, ex Bur. 

♦Parmentiera alata (Kunth) Miers 

Crescentia alata Kunth 
Spathodea campanulata Beauv. 
♦Spathodea nilotica Seem. 



27 
17 




FP 


ST 


2 








2 

2, 7. 17 

I, 2, 7, 13, 


17. 


FP 
FP 


St 


25 




FP 




2 

2. 17, 27 

15 




FP 


ST 
ST 



Sesamum radiatum Schumacher 



124. pedaliaceae 



2, 17 



FP 



125. acanthaceae 

Acanthopale decempedalis C. B. Clarke 
Acanthopale sp. 

FP : Boughey 136. 
Acanthus montanus (Nees) T. Anders. 

Adhatoda maculata C. B. Clarke 
Adhatoda orbicularis (Lindau) C. B. Clarke 
Adhatoda robusta C. B. Clarke 
Adhatoda tristis Nees 
Afrofittonia silvestris Lindau 
AsYSTASiA gangetica (L.) T. Anders. 
AsYSTASiA MACROPHYLLA (T. Anders.) Lindau 
AsYSTASiA vogeliana Benth. 
Brachystephanus longiflorus Lindau 
Brachystephanus occidentalis Lindau 
Brillantaisia lamium (Nees) Benth. 

Not recorded from FP in 17. The record from FP in 

2 may be based on Fl. West Trop. Afr., ed. i, where 

the distribution is given as 'Biafran Is.' 
Brillantaisia madagascariensis T. Anders, ex Lindau 
Brillantaisia nitens Lindau 

Not recorded from FP in 17. 
Brillantaisia cf. owariensis Beauv. 

FP : EscarrS j-6. 



2, 17 




FP 
FP 


I, 2, 7, II, 


13, 


FP 


17 






2, 17 




FP 


2. 17 




FP 


2, 17 




FP 


2, 17 




FP 


17 




FP 


I, 13. 17 




FPPST 


2, 17 




FP 


2, 17 




FP 


2, 17 




FP 


I 




ST 


I, 2, 9, II, 


13 


FP ? P ST 



17 

2 



FP 

FP 

FP 



I 






ST 


I, 2, II 


, 13 


. 17 


FP ST 


2 






FP 


17 






FP 


16 






PST 


I. 15 








2, 17 






FP 


II, 17, 


25. 


26 


FPPST 


I. 2 









OF THE GULF OF GUINEA 385 

Brillantaisia patula T. Anders. 
Brillantaisia vogeliana (Nees) Benth. 
Crossandrella dusenii (Lindau) S. Moore 

Not recorded from FP in 17. 
DiCLiPTERA LAXATA C. B. Clarke 
Dicliptera verticillata (Forssk.) Christensen 

D. umbellata (Vahl) Juss. 
DiscHiSTOCALYX THUNBERGiiFLORUS (T. Andcrs.) Benth. 
Elytraria marginata Vahl 

E. acaulis sensu Exell 
? FiLETiA AFRicANA Lindau FP 

FP : Musola, Mildbraed yooi. Not recorded from FP 

in 17. This may be Asy stasia decipiens Heine, but 

that species is also not recorded from FP. 
♦Graptophyllum pictum (L.) Nees ex Griff. i, 2, 14 FP P ST 

Heteradelphia paulowilhelmia Lindau i ST E 

Hypoestes aristata (Vahl) Solander ex Roemer & 

Schultes 17 FP 

H. insularis T. Anders. 2 

Hypoestes triflora (Forssk.) Roemer & Schultes 

'trifolia' in 2 2, 17 FP 

Hypoestes sp. i ST 

ISOGLOSSA GLANDULIFERA Lindau 2, 1 7 FP 

IsoGLOSSA sp. B 17 FP 

JusTiciA EXTENSA T. Anders. FP 

Duvernoia extensa (T. Anders.) Lindau 2 

Not recorded from FP in 17. This may be /. laxa. 
JusTiciA INSULARIS T. Anders. 2, 17 FP 

JusTiciA LAXA T. Andcrs. 2, 17 FP 

JusTiciA MANNii T. Anders. FP E ? 

The type is from FP but is apparently untraceable. 

Anderson himself named two sheets of Mann 1435 (K) 

as this but, according to E. Milne-Redhead (in litt.), 

they are definitely not the species described but are 

Adhadota orbicularis. 
JusTiciA PREUssii (Lindau) C. B. Clarke FP 

FP : Musola, Mildbraed 7047. Not recorded from FP 

in 17. 
JusTiciA TENELLA (Nees) T. Andcrs. i. 17 FP P ST 

JusTiciA THOMEENSis Lindau I ST E 

Mendoncia gilgiana (Lindau) Benoist FP 

FP : Musola, Mildbraed 6ggo ; between Musola and 

Concepcion, Wrigley & Melville 54g (K) . Not recorded 

from FP in 17. 
MiMULOPSis soLMSii Schweinf. 

M. violacea Lindau 
Nelsonia canescens (Lam.) Sprengel 
Oreacanthus mannii Benth. 
Phaulopsis falcisepala C. B. Clarke 

Phaulopsis micrantha (Benth.) C. B. Clarke i, 16, 25 ST A 

Pseuderanthemum ludovicianum (Biittner) Lindau 

Eranthemum ludovicianum Biittner 
Pseuderanthemum tunicatum (Afzel.) Milne-Redh. 
P. nigritianum (T. Anders.) Radlk. 



17 


FP 


2 




I. 17 


FPP 


2, 17 


FP 


17 


FP 


I, 16, 25 




17 


FP 


2 




17 


FP 


2 





386 



ANGIOSPERMS OF THE ISLANDS 



Rhinacanthus virens (Nees) Milne-Redh. var. virens 
R. dewevrei De "Wild. & Dur. 
This is recorded from ST in 17 but I have no evidence 
of its occurrence in that island. 
♦RuELLiA GRAECiZANS Backer 
RuELLiA PRiMULOiDES (T. Anders. ex Benth.) Heine 

Endosiphon primuloides T. Anders, ex Benth. 
RuNGiA PAXiANA (Lindau) C. B. Clarke 
ScHAUERiA POPULIFOLIA C. B. Clarke 
ScLEROCHiTON PREUSSii (Lindau) C. B. Clarke 

Not recorded from FP in 17. 
Stenandriopsis guineensis (Nees) Benoist 

Crossandra guineensis Nees 
Stenandriopsis thomensis (Milne-Redh.) Heine 

Crossandra thomensis Milne-Redh. 
♦Thunbergia alata Bojer ex Sims 
Thunbergia vogeliana Benth. 
Whitfieldia elongata (Beauv.) De Wild. & Dur. 

W. longifolia T. Anders, (incl. var. perglahra (C. B. 
Clarke) Hutch. & Dalziel) 



II, 17 

I 



15 



PST? 



2, 17 



FP 



2, 17 




2, 17 


FP 


2, 17 


FP 


2 


FP 



FP 



ST 







ST 


1. 15. 17 






I, 8 




ST 


2, 17 


FP 




17 


FP 





126. verbenaceae 

Clerodendrum bipindense Giirke 
Clerodendrum capitatum (Willd.) Schumacher 

FP : Musola, Mildbraed 7043. Not recorded from 
FP in 17. 
Clerodendrum dusenii Giirke 
C. thomasii Moldenke 
Premna macrosiphon Baker 
Not recorded from FP in 17. 
Clerodendrum globuliflorum B. Thomas 
Clerodendrum grandifolium Giirke 

FP : Mildbraed 6345. Not in 17. 
♦Clerodendrum japonicum (Thunb.) Sweet 

C. fragrans var. pleniflorum Schauer 
Clerodendrum melanocrater Giirke 
Clerodendrum silvanum Henriq. 

C. buchholzii Giirke 
♦Clerodendrum speciosissimum Van Geert ex Morren 
C. fallax Lindl. 
Not listed in 2 but see photograph following p. 352. 
Clerodendrum splendens G. Don 

Not recorded from FP in 17. 
Clerodendrum thomsoniae Balf. f. 

C. scandens sensu Exell 
Clerodendrum umbellatum Poiret 

C. scandens Beauv. 
Clerodendrum volubile Beauv. 

♦DURANTA REPENS L. 

♦Lantana camara L. 
Premna angolensis Giirke 
Premna milnei Baker 
Stachytarpheta indica (L.) Vahl 



2, 17 



17 

I 
2 



2, 7 



FP 
FP 



FP ST 



17 


FP 
FP 


ST 


II 

17 

1 


FP 
FP 


ST A 


17 

I, 11 


FP 


ST 


2 







FP 



II 




ST 


I 






17 




FP 


2 






I, 17 




FPP 


I. II 




ST 


I. 13. 


25 


§T 


II. 25 




ST 


2. 17 




FP 


17 




FP 



OF THE GULF OF GUINEA 



387 



ViTEX LEHMBACHii Gurke 
FP : Bokoko, Burchardt. 
17- 



FP 



Not recorded from FP in 



ViTEX MiLNEi Pieper 

V. cienkowskii sensu Guinea 


17 

2 


FP 




127. AVICENNIACEAE 








AVICENNIA GERMINANS (L.) L. 

A. africana Beauv. 
A. nitida J acq. 


14, 15, 16 

1. 13. 17 

2, 7 


FP 


ST A 


128. LABIATAE 








ACHYROSPERMUM OBLONGIFOLIUM Baker 

COLEUS aff. PREUssii Giirke 


I, 2, 17 


FP 
FP 


ST 



E? 



FP : Moka, Mildbraed 7103. Coleus preussii is not 

mentioned in 17 although the syntypes are from Buea. 

It is presumably a species of Plectranthus. 
*' Coleus tuber osus' 

Perhaps C. parviflorus Benth. 
Elsholtzia sp. 

HoMALOCHEiLOS RAMOsissiMus (Hook. f.) J. K. Morton 
Plectranthus ramosissimus Hook. f. 

HOSLUNDIA OPPOSITA Vahl 

H. verticillata Vahl 
Hyptis lanceolata Poiret 
Leonotis nepetifolia (L.) Alton f. 

Not recorded from FP in 17. 
Leucas deflexa Hook. f. 
Leucas martinicensis (Jacq.) Alton f. 
♦Mentha rotundifolia (L.) Huds. 

*OciMUM BASILICUM L. 

OciMUM CANUM Sims 

♦OCIMUM GRATISSIMUM L. 

O. viride Willd. 

Platostoma africanum Beauv. 
Plectranthus assurgens (Baker) J. K. Morton 
Plectranthus decurrens (Giirke) J. K. Morton 

Coleus decurrens Giirke 
Plectranthus esculentus N. E. Brown 
Coleus dazo A. Chev. 
Not recorded from FP in 17. 
Plectranthus glandulosus Hook. f. 
Plectranthus insignis Hook. f. 

FP : Melville & Wrigley ^Sg (K) ; Mildbraed 7143. 
Not recorded from FP in 17. 
Plectranthus punctatus L'Herit. subsp. punctatus 
Pycnostachys meyeri Giirke 

P. volkensii Giirke 
♦Salvia coccinea Buc'hoz ex Etlinger 
Satureja pseudosimensis Brenan 
Calamintha simensis sensu Guinea 



FP 



I 






P? 


17 






FP 


2 








9, II, 17 






FP A 


2 








17 






FP 


I, 2, II, 


13. 


17. 


FP P ST A 


25, 26 








17 






FP 


17 






FP 


13 






ST 


I, 2, 13, 


16, 


17, 


FP ST A 


25 








2, 17 






FP 


17 






FP P ST A 


I, 2, II, 


13. 


25. 




26 








I. 2, 17 






FP ST 


17 






FP 


17 






FP 


2 









2, 17 



17 
17 

2 

I 

17 

2 



FP 



FP 
FP 



FP 
FP 



FP 



ST 



I, 2, l8 


FPP 


i8 


FP 


2, l8 


FP 


2 




2, l8 


FP 


2 


FP 



388 ANGIOSPERMS OF THE ISLANDS 

SoLENOSTEMON MANNii (Hook. f.) Baker 17 FP 

SoLENOSTEMON MONOSTACHYus (Beauv.) Briq. subsp. 

MONOSTACHYUS I, II, I3, I7, 25 FP P ST A 

SoLENOSTEMON REPENS (Giirke) J. K. Morton 17 FP 

Stachys aculeolata Hook. f. 2, 17 FP 

♦StACHYS ARVENSIS (L.) L. II ST 

MONOCOTYLEDONES 

129. HYDROCHARITACEAE 

Ottelia ulvifolia (Planchon) Walp. 18 FP 

130. POTAMOGETONACEAE 

POTAMOGETON SCHEWEINFURTHII A. Benn. l8 FP ? 

131. COMMELINACEAE 

Aneilema beniniense (Beauv.) Kunth 

Aneilema dispersum Brenan 

Aneilema umbrosum (Vahl) Kunth subsp. umbrosum 

A. nigritianum (C. B. Clarke) Hutch. 
Buforrestia mannii C. B. Clarke 
' Commelina africana C. B. Clarke' 

Possibly C. africana L. var. mannii (C. B. Clarke) 
Brenan is intended. This is not recorded from FP 
in 18. 
Commelina benghalensis L. 2, 16 FP 

Not recorded from FP in 18. 
Commelina cameroonensis J. K. Morton 
Commelina capitata Benth. 
Commelina congesta C. B. Clarke 

C. condensata C. B. Clarke 
Commelina diffusa Burm. f. subsp. diffusa 

C. nudiflora L. pro parte 
Commelina diffusa subsp. Montana J. K. Morton 
Commelina erecta L. subsp. erecta 

C. vogelii C. B. Clarke 
Cyanotis barbata D. Don 

C. mannii C. B. Clarke 
Palisota barteri Hook. 
? Palisota bicolor Mast. 
Palisota bracteosa C. B. Clarke 
Palisota hirsuta (Thunb.) K. Schum. 

P. thyrsiflora Benth. 
Palisota mannii C. B. Clarke 
Palisota pedicellata K. Schum. 
Palisota preussiana K. Schum. ex C. B. Clarke 
PoLLiA condensata C. B. Clarke 
PoLLiA MANNII C. B. Clarke 

POLYSPATHA PANICULATA Benth. 

p. glaucescens (C. B. Clarke) Hutch. 
♦Rhoeo spathacea (Swartz) Stearn FP ST ? 

R. discolor (L'Herit.) Hance i, 2 

Stanfieldiella brachycarpa var. hirsuta (Brenan) 

Brenan 18 FP ? 



18 


FP 


2, 18 


FP 


18 


FP ST 


I, 2, 25 




I, 8, 13, 18 


FP P ST A 

FP t 


2. 25 


18 


18 


FP ] 


2 




18 


FP 


2 




2, 18 


FP 


2, 18 


FP 


I, 2, 18 


ST 1 


18 


FP t 


2 




2, 18 


FP ; 


I. 13 


P ST A E i 


18 


FP 


I, 2, 8, 18 


FP P ST J 


I, 2, 18 


ST "! 


2, 7, 18 


FP 


2 





OF THE GULF OF GUINEA 389 

Stanfieldiella imperforata (C. B. Clarke) Brenan 16 FP ? P ST 

Buforrestia imperforata C. B. Clarke i, 2, 13 

Brenan (in Kew Bull. 14 : 285 (i960)) states that one 

of the syntypes is from FP, but he considers that it 

may be S. brachycarpa var. hirsuta (see above). 

♦Zebrina pendula Schnizl, i. 13 P ST 

132. flagellariaceae 

Flagellaria guineensis Schumacher 2 FP 

Not recorded from FP in 18, 

133. eriocaulaceae 

Eriocaulon zambesiense Ruhl. 18 FP 

Mesanthemum radicans (Benth.) F. Komicke 2, 18 FP 

134. BROMELIACEAE 

*Ananas comosus (L.) Merr. i FP P ST A 

A. sativus Lindl. 2 

135. MUSACEAE 

*MusA CHiNENSis Swect 2 FP 

*MUSA PARADISIACA L. I, 2, I3, 25 FP P ST A 

♦MusA PARADISIACA var. SAPiENTUM (L.) Kuntze 2 FP ST 

M. sapientum L. 13 

M. paradisiaca subsp. sapientum (L.) K. Schum. 25 
*MusA PARADISIACA var. viTTATA (W. Ackerm. ex 

Rodigas) K. Schum. I ST 

♦MusA ROSACEA Jacq. I ST 

*MusA TEXTiLis N6e 2 FP 

•MuSA VENTRICOSA Welw. 1 3 ST 

♦Raven ALA madagascariensis Sonnerat ^ FP 

*Strelitzia reginae Banks 2 FP 

136. zingiberaceae 

Aframomum chlamydanthum Loes. & Mildbr. 18 FP 

Aframomum daniellii (Hook, f.) K. Schum. i, 2, 8, 9, 11, 13, FP P ST 

18, 25, 26 
Aframomum keniense R. E. Fries FP 

FP : Melville &- Wrigley 487 (K.) Not in 18. 

Aframomum limbatum (Oliver & Hanb.) K. Schum. 2, 7, 18 FP 

Aframomum melegueta (Roscoe) K. Schum. i, 2, 18 FP P ST 

Aframomum pilosum (Oliver & Hanb.) K. Schum. 2, 18 FP 

Aframomum sceptrum (Oliver & Hanb.) K. Schum. FP 

FP : Melville &> Wrigley 555 (K). Not recorded from 

FP in 18. 

CosTus AFER Ker-Gawl. i, 2. 11, i8, 25, FP P ST 

26 

CosTus ENGLERANUS K. Schum. l8 FP 

CosTus GiGANTEUS Welw. ex Ridley i, 11, 13, 25 P ST A 

CosTus LUCANUSIANUS J. Braun & K. Schum. 2, 18 FP 

CosTus SPECTABILIS (Feuzl) K. Schum. I ST 

♦Curcuma domestica Valeton i P 

♦Elettaria repens (Sonnerat) Baillon X ST ? 



390 



ANGIOSPERMS OF THE ISLANDS 



♦Hedychium coronarium Koenig 
Renealmia africana Benth. & Hook. 
Renealmia cincinnata (K. Schum.) T. Dur. 
Renealmia grandiflora Baker 

Alpinia africana Ridley 
Renealmia macrocolea K. Schum. 
Probably the same as R. africana. 
Renealmia mannii Hook. f. 
*Zerumbet speciosum Wendl. 

Alpinia speciosa (Wendl.) K. Schum. 
♦Zingiber officinale Roscoe 
♦Zingiber zerumet (L.) Smith 

137. cannaceae 

♦Canna indica L. 

This should perhaps be treated as a distinct African 
species, C. hidentata Bertol. 

138. marantaceae 

Halopegia azurea (K. Schum.) K. Schum. 

Not recorded from FP in 18. 
' Hybophrynium mannianum ' 

I cannot trace this name. Perhaps it is an error for 

H. braunianum (see Trachyphrynium below). 
♦Maranta arundinacea L. 
Marantochloa filipes (Benth.) Hutch. 
Clinogyne filipes (Benth.) K. Schum. 
? Marantochloa flexuosa (Benth.) Hutch. 
Clinogyne flexuosa (Benth.) K. Schum. 

This species was recorded from FP in Fl. W. Trop. Afr.. 

ed. I, but in 18 it was divided between M. leucantha 

(K. Schum.) Milne-Redh. and M. purpurea (Ridley) 

Milne-Redh. and neither of these species was recorded 

from FP. 
Marantochloa leucantha (K. Schum.) Milne-Redh. 

FP : Melville <S- Wrigley 554 (K). Not recorded from 

FP in 18. 
Marantochloa mannii (Benth.) Milne-Redh. 

Phrynium mannii (Benth.) K. Schum. 
Marantochloa ramosissima (Benth.) Hutch. 

Clinogyne ramosissima (Benth.) K. Schum. 
Sarcophrynium brachystachys (Benth.) K. Schum. 
Phrynium brachystachys (Benth.) Kornicke 

Not recorded from FP in 18. 
Thaumatococcus daniellii (Benn.) Benth. 

Not recorded from FP in 18. 
Trachyphrynium braunianum (K. Schum.) Baker 



1, 2, 

2. 18 

I 
I 


II 




FPP 
FP 
P 
ST 


-,2 
2 






FP 


2, 18 






FP 

ST 


13 
I, 2, 
2 


13. 


25 


FPP ST 
FP 


I, II, 


13 




P? ST A 



X 

18 

2 



18 
2 
18 
2 



I, 2, 7, 9, 18 



FP 
FP 



PST A 



FP 

FP? 



FP 

FP 
FP 

FP 

FPP ST 
FP 



139. LILIACEAE 



♦Asparagus officinalis L. 
Chlorophytum sparsiflorum Baker 
Gloriosa simplex L. 

Not recorded from FP in 18. 



13 

2, i{ 
2 



ST 



FP 
FP 



OF THE GULF OF GUINEA 



391 



Gloriosa superba L. 

Not recorded from FP in 18. 
WuRMBEA TENUIS (Hook. f.) Baker 

Smilax kraussiana Meissner 



140. smilacaceae 



141. araceae 



Amorphophallus mannii N. E. Brown 
Anchomanes difformis var. pallidus (Hook.) Hepper 
A. petiolatus (Hook.) Hutch. 

The plant recorded as A. difformis (Blume) Engl, in 2 

is likely to be this variety. 
Anubias barteri Schott 
Anubias heterophylla Engl. 

Not in 18. 
♦Caladium bicolor (Alton) Vent. 
Cercestis stigmaticus N. E. Brown 
C. dinklagei Engl, pro parte 

♦COLOCASIA ESCULENT A (L.) Schott 

CuLCASiA ANGOLENSis Welw. ex Schott 
CuLCASiA LANCiFOLiA N. E. Brown 

C. insulana N. E. Brown 
CuLCASiA parviflora N. E. Brown 
CuLCASiA SCANDENS Beauv. 
Cyrtosperma senegalense (Schott) Engl. 
Nephthytis constricta N. E. Brown 
Nephthytis poissonii (Engl.) N. E. Brown 

PiSTIA STRATIOTES L. 

Not recorded from FP in 18. 
Rhaphidophora africana N. E. Brown 

Afrorhaphidophora africana (N. E. Brown) Engl. 
Rhektophyllum MiRABiLE N. E. Brown 
Sauromatum venosum (Alton) Kunth 

J aimenostea fernandopoana Guinea & Moreno 
Stylochiton zenkeri Engl. 
*? Xanthosma sagittifolium (L.) Schott 

This may be X. mafaffa Schott. 
♦Xanthosma violaceum Schott 



2 

2, 18 

18 



FP 
FP 

FP 



2. 18 


FP 




E 


2 ?, 18 


FP 




E 


2, 7 








2, 18 


FP 






2 


FP 






I, 2, 13 


FPPST 




2, 18 


FP 






2 








I, 2, 13 


FPPST 




I. 2, 7, 18 


FP 


ST 




18 


FP 






2 








2, 18 


FP 






I 




ST 




2, 18 


FP 






2, 18 


FP 






18 


FP 






2 


FP 






18 


FP 






2 








2. 7, 18 


FP 






18 


FP 






2 








2. 18 


FP 






13, 16 




ST A 





FP 



142. lemnaceae 



Lemna perpusilla Torrey 



143. amaryllidaceae 

♦Allium cepa L. 

♦Allium sativum L. 

♦Crinum asiaticum L. 

Crinum jagus (Thompson) Dandy 
C. giganteum Andr. 
Not recorded from FP in 18, presumably in error, as 
it is recorded in Fl. W. Trop. Afr., ed. i, under C. 
giganteum, on the basis of Exell y68, the sheet of which 
in BM bears a ticket reading ' Seen for Revised Edition 
of F.W.T.A.' 



15 



ST 



2. 13 


FP 


ST 


13 




ST 


I 




ST? 


I 


FP 


ST 



392 ANGIOSPERMS OF THE ISLANDS 

Crinum natans Baker 2, i8 FP 

Haemanthus cinnabarinus Decaisne 2, i8 FP 

Haemanthus multiflorus Martyn 18 FP 

♦HipPEASTRUM PUNicEUM (Lam.) Urban i P ST 

Hymenocallis littoralis (Jacq.) Salisb. 16 A 

144. iridaceae 

ACIDANTHERA DIVINA Vaupel 2, l8 FP 

♦Neomarica caerulea (Lodd.) Sprague i, ii FP P ST 

Marica sahinii Lindl. 2 

145. dioscoreaceae 

♦Dioscorea alata L. I, II, 13, 16, 25 ST A 

DiOSCOREA BULBIFERA L. I, 2, I3 FP ST 

Not recorded from FP in i8. 

♦Dioscorea cayenensis Lam. i, 9, 11, 13, 16 P ? ST A 

Dioscorea dumetorum (Kunth) Pax i, 2, 11, 13, 16, FP ST A 

25, 26 
Not recorded from FP in 18. 

Dioscorea elephantipes (L'H^rit.) Engl. FP 

D. elephantopus Sprengel 2 

Dioscorea minutiflora Engl. i, 2, 13 FP ? ST 

The record from FP in 2 is probably based on a mis- 
identification of D. smilacifolia. 

Dioscorea preussii Pax 2 FP 

Not record